Document 14737023

advertisement
CROSS-DOMAIN DETERRENCE
IN U.S.–CHINA Strategy
A Workshop Report
James Scouras
Edward Smyth
Thomas Mahnken
Copyright 2014 The Johns Hopkins University Applied Physics Laboratory LLC.
All rights reserved.
Inquiries
Questions, comments, and requests for permission to make copies of any part of this document
should be mailed to:
Dr. James Scouras
National Security Analysis Department
The Johns Hopkins University Applied Physics Laboratory
11100 Johns Hopkins Road
Laurel, Maryland 20723
Disclaimer
The views expressed in this report should not be construed as the views of any of the
organizations with which the authors or workshop participants are affiliated.
Library of Congress Cataloging-in-Publication Data
Scouras, James, author.
Cross-domain deterrence in U.S.-China strategy : a workshop report / James Scouras, Edward
Smyth, Thomas Mahnken.
pages cm
1. Deterrence (Strategy) 2. United States--Foreign relations--China. 3. China--Foreign relations-United States. I. Smyth, Edward (Edward A.) author. II. Mahnken, Thomas G., 1965- , author.
III. Johns Hopkins University. Applied Physics Laboratory, issuing body. IV. Title.
U162.6.S36 2014
355’.033573--dc23
2014012446
CONTENTS
Preface................................................................................................................................... v
Summary.............................................................................................................................. vii
1.
Workshop Introduction.................................................................................................. 1
2.
Chinese Perspectives on Deterrence.............................................................................. 9
A. Presentation Summary............................................................................................. 9
B. Key Questions and Participants’ Responses............................................................ 13
3.
Deterring Chinese Cyber Attacks................................................................................. 19
A. Presentation Summary........................................................................................... 19
B. Key Questions and Participants’ Responses............................................................ 22
4.
Deterring Chinese Attacks on U.S. Space Capabilities................................................. 27
A. Presentation Summary........................................................................................... 27
B. Key Questions and Participants’ Responses............................................................ 31
5.
Deterring Chinese Nuclear Use................................................................................... 37
A. Presentation Summary........................................................................................... 37
B. Key Questions and Participants’ Responses............................................................ 40
6.
Future Research........................................................................................................... 45
7.
Final Thoughts.............................................................................................................. 51
PREFACE
This study had its origin in a discussion
between the Director of the Johns Hopkins
University Applied Physics Laboratory
(JHU/APL), Dr. Ralph Semmel, and the former
Deputy Assistant Secretary of Defense for
Nuclear and Missile Defense Policy, Dr. Bradley
Roberts, in June 2012. During that discussion,
Dr. Roberts indicated he would be interested in
JHU/APL-suggested approaches to thinking
about cross-domain deterrence.
Consequently, JHU/APL initiated an
internally funded study to address crossdomain deterrence. The Laboratory allocates
a portion of its internal funding to a program
of innovative research with the potential
for significant impact on critical national
challenges. This study, funded under that
program, focuses on clarifying cross-domain
deterrence issues and identifying productive
research approaches. The first stage of the
study culminated in a briefing provided to
Dr. Roberts and his staff in January 2013.
The Cross-Domain Deterrence Workshop,
documented by this report, concludes the
second stage of JHU/APL’s research.
The study leaders are James Scouras
(Principal Investigator), Edward Smyth,
and Thomas Mahnken. Dr. Scouras is
a National Security Studies Fellow at
JHU/APL. Mr. Smyth is the Branch
Supervisor of JHU/APL’s National Security
Analysis Department’s Analysis, Modeling,
and Simulation Branch. Dr. Mahnken is
a Senior Research Professor of Strategic
Studies at the Johns Hopkins University
School of Advanced International Studies.
Other
JHU/APL
researchers—notably
Dr. Antonio DeSimone, Dr. Danielle Wood,
and Mr. Michael Shehan—contributed
cyber and space domain expertise to
the study.
We are grateful to the experts who
participated in the workshop, especially to
the presenters and discussion facilitators, all
listed in Table 1 of this report. In addition,
we thank Margaret Harlow for taking
notes and developing draft summaries of
workshop sessions and Deborah Schlichting
for managing the workshop logistics.
As the primary documentation of the
JHU/APL Cross-Domain Deterrence Work­
shop, this report should be of interest to policy
makers, analysts, and citizens concerned with
emerging deterrence challenges, especially
vis-à-vis China.
v
SUMMARY
This report documents the Cross-Domain
Deterrence Workshop
conducted
on
26 June 2013 at the Johns Hopkins University
Applied Physics Laboratory (JHU/APL) in
Laurel, Maryland. The overarching objectives
of the workshop were to identify: (1) the
challenges in deterring actions in one domain
of warfare (e.g., cyber, space, nuclear,
conventional, etc.) by posing retaliatory threats
in another domain and (2) research needs and
promising research approaches for advancing
our understanding of these challenges and
forging effective cross-domain policies and
strategies.
For purposes of this workshop, we defined
domains as categories of weapons effects—
nuclear, conventional, space, cyber, missile
defenses, chemical, biological, etc. Crossdomain deterrence, therefore, involves making
retaliatory threats from one domain to prevent
attacks from another. Because the topic of
cross-domain deterrence is so broad,
we limited our scope by both actors and
domains under consideration. In particular,
this workshop focused on China, which
arguably poses the broadest array of
cross-domain deterrence challenges. We
further limited ourselves primarily to the
nuclear, cyber, space, and conventional
military domains.
The workshop sessions were Chinese
Perspectives on Deterrence; Deterring Chinese
Cyber Attacks; Deterring Chinese Attacks on
U.S. Space Capabilities; Deterring Nuclear
Threats in a U.S.–China Crisis or Conflict; and
Future Research. In this summary, we present
selected insights derived from workshop
discussions, not session by session, but rather
according to the following cross-cutting
categories.
Deficiencies
• Available Chinese sources related to
cross-domain issues are insufficient to
confidently anticipate Chinese actions in
a crisis or conflict.
• The United States lacks clear, effective,
and commonly understood cross-domain
strategies, especially for deterring and
responding to cyber and space attacks.
Controversies
• Some participants argued that Chinese
leadership sees China’s nuclear weapons
as useful for deterring U.S. retaliatory
actions in several domains of warfare.
Other participants argued that China
sees its nuclear weapons as principally
limited to deterring others’ nuclear threats
or use and conventional attacks highly
threatening to Chinese Communist Party
rule.
• Most participants believed it was
important
to
develop
a
deep
understanding of China’s history,
strategic culture, leadership, doctrine,
and decision-making processes. Others
believed that we need only understand
China’s geopolitical circumstances and
logically analyze its options for actions
in this context.
Needs
• The United States needs to place greater
emphasis on denying benefits to both
space and cyber attacks.
• The United States should try to strengthen
international resolve against kinetic
vii
attacks in space and destructive cyber
attacks.
Future Research
Future research topics include additional
areas we need to understand about China and
scenarios worthy of further analysis.
Understanding China. Several participants focused on understanding central
aspects of China’s strategic culture—its role
in the international system, its values, and
its fears. Other participants focused more
directly on the need to understand China’s
approach to nuclear warfare and deterrence, deterrence of the strong by the weak,
separating peacetime deterrence of the use
of force and deterrence of further escalation during wartime or limited conflict, and
coercive persuasion.
Scenarios. Many participants suggested
scenarios involving combinations of actors
in addition to those involved in a direct
confrontation between the United States and
China. Others suggested scenarios in which
the internal situation in China is an important
driver, scenarios involving a cyber attack
that causes significant unintended damage,
scenarios exploring reactions to nuclear
use, and scenarios designed for developing
and assessing options for responding to
viii
various circumstances. Several participants
emphasized the importance of more
realistically representing Chinese doctrine
and decision making in scenario analysis.
Final Thoughts
In reflecting on the workshop discussions,
we offer the following additional suggestions
for avenues of future research. First, we
might draw lessons that could be applied
to emerging threats and future scenarios
from studying the historical applications
of deterrence in those situations in which
multiple domains were in play. Second,
we might also benefit from comparative
analyses of other states’ perspectives and
policies—particularly those of China and
Russia—regarding deterrence, cross-domain
and otherwise. Finally, our workshop was
conducted at what can be characterized
as a conceptual level. However, it became
clear that considering cross-domain issues
at an abstract level is limiting as almost
all cross-domain deterrence decisions are
context dependent. Thus, scenario analysis
must be an integral component of future
research. Moreover, as several participants
noted, scenario analysis could greatly benefit
from the analysts trying to emulate Chinese
decision making.
1
WORKSHOP INTRODUCTION
This report documents the Cross-Domain
Deterrence Workshop conducted on
26 June 2013 at the Johns Hopkins
University Applied Physics Laboratory
(JHU/APL) in Laurel, Maryland. In this chapter,
we present the workshop objectives, discuss
our approach to scoping workshop topics,
clarify our usage of the terms “domain” and
“deterrence,” and summarize the workshop
read-ahead papers. We then present the
workshop agenda, list workshop participants
and their affiliations, and describe our
approach to developing this report.
Objectives
The overarching objectives of the workshop
were to identify: (1) the challenges in deterring
actions in one domain of warfare (e.g., cyber,
space, nuclear, conventional, etc.) by posing
retaliatory threats in another domain; and
(2) research needs and promising research
approaches for advancing our understanding
of these challenges and forging effective
cross-domain policies and strategies.
Scope and Terminology
Because the topic of cross-domain deterrence
is so broad, we limited our scope by both
actors and domains under consideration.
In addition, to facilitate dialogue, we
established a definition of deterrence for use
throughout the workshop.
Actors. Although a variety of actors
could pose threats for which cross-domain
deterrence may be useful, this workshop
focused on China, which arguably poses the
broadest array of potential threats for which
cross-domain deterrence challenges appear
most problematic. In particular, both the
cyber and space challenges from China are
relatively new and growing, with ill-defined
norms of international behavior and poor
visibility into China’s strategic thinking and
decision making.
In focusing on China, we recognize that
cross-domain challenges from other states
(and non-state entities) are not lesser included
cases. That is, we should not expect that what
we learn about China will necessarily apply
to Russia, Iran, or others. These states need
to be addressed as well, and comparative
analysis of the various bilateral deterrence
relationships might prove very informative.
We also do not consider non-state actors.
These pose unique deterrence challenges that
require an analytic approach that is different
from that adopted for this workshop.
Domains. For purposes of this workshop,
we have defined domains as categories of
weapons effects—nuclear, conventional,
space, cyber, missile defense, electronic,
chemical, biological, etc. Of these domains,
space is somewhat of an outlier, as it is more
commonly thought of as a place in which
actions occur (the other domains of that
ilk being land, air, sea, and cyberspace).
However, one can think of weapons (both
those based in space and elsewhere) that
attack space capabilities (again, those based
in space and elsewhere) as consistent with
a domain characterized by a weapon effect.
1
Workshop Introduction
In any event, we recognize that continuing
ambiguity in the term “domain” remains an
impediment to clear thinking and effective
dialogue, but addressing terminology was not
a focus of this workshop.
Also, we focused on military domains,
rather than a whole-of-government approach
to deterrence that would include the full spectrum of levers of national influence including
trade, aid, diplomacy, etc. We emphasized the
nuclear, cyber, space, and conventional military domains. Other important military domains, such as air and missile defenses, were
only tangentially discussed. Similarly, military
domains open to some adversaries but not to
the United States (e.g., biological, chemical)
were not addressed.
These choices were based on our judgments
that the most challenging cross-domain issues
arise in cyber and space and that isolating
the nuclear domain remains a critical goal.
Moreover, in thinking through cross-domain
options, the conventional domain also looms
large, given U.S. superiority in that domain
and its ability to scale effects. Non-military
domains are important, perhaps even central,
to deter some cyber and space attacks, but
their utility appears to be more relevant to
threats at the lower end of the spectrum of
violence.
Deterrence. For purposes of this
workshop, we defined “deterrence” as the
strategy that seeks to prevent actions by the
specter of retaliation. That is, deterrence is
only about sticks, rather than both carrots
and sticks. Also, deterrence is about
the anticipation of punishment rather
than making actions more difficult to
accomplish or mitigating their con­
sequences. Notwithstanding other usages
of this term, our usage is consistent with most
dictionary definitions, including that of the
Department of Defense.1
Cross-domain deterrence, therefore,
involves making retaliatory threats from one
domain to prevent attacks from another. Denial
strategies also have their role in preventing
unwanted actions, and, indeed, we explicitly
addressed their utility in a number of questions
posed during the workshop. However, for
the sake of clarity, we do not define them as
deterrence.
Finally, the workshop was conducted
at the unclassified level, and—other than
presentations—all comments were treated
on a not-for-attribution basis. Presentation
authors are identified on the workshop
agenda, and presentation summaries in this
report are associated with presenters.
Read-Ahead Papers
Two papers were provided to participants
prior to the workshop. These were selected to
ensure that all participants had a foundation
of knowledge on cross-domain terminology
and challenges and Chinese perspectives on
deterrence. These papers, both by authors
who also participated in the workshop, are
summarized here:
Dean Cheng, “Chinese Views on
Deterrence,” Joint Forces Quarterly,
1st Quarter, 2011. Mr. Cheng focuses on
Chinese definitions related to deterrence
and Chinese operationalization of deterrence
theory, which differ in important ways
from U.S. definitions and practices,
notwithstanding
some
fundamental
similarities.
1
Department of Defense, Dictionary of Military and Associated Terms, Joint Publication 1-02,
8 November 2010, as amended through 15 June 2013.
2
Cross-Domain Deterrence in U.S.–China Strategy
In terms of similarities, the core tenet that
deterrence is based on the credible threat
of use of military power is central to both
Chinese and U.S. theories of deterrence.
Moreover, both sides recognize that credible
threats require both military capability and
the will to use that capability as judged by
the side subjected to deterrent threats.
In China’s application of deterrence
theory, military capabilities include not only
the traditional nuclear and conventional
forces but also, increasingly, space and
information capabilities. The roles of all
these capabilities are evolving as technology
advances:
• China distinguishes various levels of
nuclear deterrence. Its strategy thus far
has been one of “minimum” nuclear
deterrence, in which a small number of
nuclear weapons can retaliate against
cities, but China may be edging toward
“moderate” nuclear deterrence, which
threatens a greater level of retaliation.
• Conventional long-range precision strike
capabilities are increasing the importance
of conventional deterrence.
• Space systems both support nuclear
and conventional deterrence and can
help neutralize an opponent’s nuclear
deterrent.
capabilities
• Information
support
deterrence in their abilities to affect
combat operations and influence
leadership and public opinion.
While military capabilities are central to
both U.S. and China’s theories of deterrence,
China has a more expansive perspective on
the capabilities that underwrite deterrence.
In China’s view, deterrence is based on “all
the components of ‘comprehensive national
power,’ including military forces, economic
power, diplomatic capabilities, and even
political and cultural unity.”
Another notable difference is that the
Chinese term for deterrence, weishe, does
not distinguish deterrence (a strategy for
preventing an unwanted action by using
threats of retaliation) and compellence (a
strategy for motivating adversary action,
again through threats of harm), whereas U.S.
deterrence theorists do. Weishe embodies
both concepts as mechanisms for compelling
an opponent to submit to the will of the
deterrer.
Yet another significant difference arises in
views of the relationship between deterrence
and warfighting. In China, these concepts are
seen as complementary. That is, deterrence
extends into the combat phase of conflict
to undermine the enemy’s will to resist. By
contrast, in the United States, the concepts
are more separated—war is the consequence
of deterrence failing.
Understanding these and other major
differences in Chinese and U.S. perspectives
on deterrence can be enhanced by
appreciation of their different histories and
geopolitical situations. Among the most
important of these factors is the lack of a
significant surprise attack experience in
China, analogous to Pearl Harbor; the lower
level of concern in China with the lessons
from World War I regarding inadvertent war;
and the greater Chinese focus on multilateral
deterrence as a result of threats from many
states on its borders in contrast to the U.S.
Note that Manzo’s definition of “domain” is not the same as that used in this workshop, and thus
neither is his definition of “cross-domain.” The existing multiplicity of definitions is a clear indicator of
the lack of a shared understanding of cross-domain issues among the U.S. strategic community.
2
3
Workshop Introduction
bilateral deterrent focus as a result of its Cold
War experience.
Vincent Manzo, “Deterrence and
Escalation in Cross-Domain Operations:
Where Do Space and Cyberspace Fit?”
Strategic
Forum,
National
Defense
University, December 2011. Although this
paper was published about 2 years ago,
it is one of the first to provide a comprehensive
overview
of
emerging
cross-domain
deterrence issues, including alternative
definitions of the term “cross-domain”
and the applicability of concepts such as
deterrence, proportionality, and escalation.
Cross-domain challenges are illustrated using
examples that focus on the roles of space and
cyberspace in potential conflicts with China.
Mr. Manzo offers two alternative
definitions of the term “cross-domain.”2 In
the first, the attacking platform uses a domain
(air, land, sea, space, or cyberspace) that
is different from that of the target platform
(e.g., a land-based missile attacking a ship).
In the second, the domain of the target is
different from the domain of the intended
consequences (e.g., attacking a satellite in
space with intended consequences for sea, air,
or land operations). By either definition, crossdomain deterrence has been commonplace
in U.S. strategy. However, increasing and
asymmetric U.S. dependencies on the space
and cyber domains, coupled with new threats
that exploit U.S. vulnerabilities in these
domains, have brought a renewed attention
to cross-domain deterrence.
Long-standing principles of deterrence
theory hold that the credibility of a
retaliatory threat is enhanced if it is logically
connected to, and has consequences that are
proportionate to, the action being deterred.
Currently, however, states lack a shared
4
framework for how these principles apply to
attacks in the space and cyber domains. This
undermines deterrence and increases the
potential for misinterpretation that could lead
to unintended escalation. Establishing such a
framework is hampered by a lack of conflict
experience in cyber and space relative
to other domains, as well as differences
in strategic cultures and asymmetries in
objectives, strategies, forces, strengths, and
vulnerabilities.
Manzo notes that even the internal U.S.
strategic community suffers from a lack of
a common understanding, which hampers
cross-domain deterrence and contingency
planning. A necessary first step toward
developing an inter state framework is that
the U.S. strategic community engage in a
comprehensive discussion of these issues,
with the objective of developing a consensus
that would facilitate integrating actions in
space and cyberspace with those of the more
traditional domains.
As a starting point for a shared
international framework, Manzo offers the
principle that, rather than focusing on the
extension of conflict from cyber and/or space
to the terrestrial domains as necessarily
escalatory, “the real-world effects of such
attacks, within the domain of the attack and
in other domains, should determine whether
they are escalatory and which responses
would be appropriate.” Among a myriad of
other considerations that come into play,
Manzo discusses the balance between
offense and defense in various domains;
the ability to substitute terrestrial for space
capabilities; the international context of
peace, crisis, or war; and the potential role
and risks of ambiguity in U.S. cross-domain
deterrence strategy.
Cross-Domain Deterrence in U.S.–China Strategy
Agenda
The workshop was conducted at an unclassified level and on a not-for-attribution basis.
The agenda appears below.
Cross-Domain Deterrence Workshop Agenda—26 June 2013
9:00 – 9:30
Workshop Introduction – James Scouras
9:30 – 11:00
Chinese Perspectives on Deterrence – Presentation by Thomas Mahnken
Discussion Facilitator – Edward Smyth
1. How much confidence can the United States have in its understanding of Chinese perspectives on deterrence, based on existing literature?
2. Is it an important issue that the United States and China have similar, yet different, definitions of deterrence?
3. What misperceptions exist in China and the United States about each other’s deterrence strategy?
4. What do the Chinese think of the U.S. discussion of “cross-
domain deterrence”?
5. China considers strategic deterrence to involve all elements of national power in a seemingly progressive pattern. How does this contrast with the U.S. process?
6. People’s Liberation Army doctrine consists of both “deterrence campaigns” as well as “warfighting campaigns.” Is this in contrast to U.S. doctrine?
11:00 – 11:15 Break
11:15 – 12:30 Deterring Chinese Cyber Attacks – Presentation by Stephen Blank
Discussion Facilitator – Antonio DeSimone
7.
Can the United States effectively dissuade cyber attacks without
demonstrating willingness to retaliate (i.e., are protection strategies and establishing behavioral norms sufficient for preventing cyber attacks)?
8. Are there effective retaliatory threats that stay within the cyber domain, or must retaliatory threats cross domains to be effective?
9.
What principles should guide the development of U.S. crossdomain deterrence policies for cyber (e.g., are red lines useful, and, if so, what should they be and how should they be communicated)?
12:30 – 1:00 Lunch
5
Workshop Introduction
1:00 – 2:15
Deterring Chinese Attacks on U.S. Space Capabilities – Presentation by Forrest Morgan
Discussion Facilitator – Alexander Ihde
10.
Can the United States effectively dissuade space attacks without demonstrating willingness to retaliate (i.e., are protection strategies and establishing behavioral norms sufficient for preventing space attacks)?
11. Are there effective retaliatory threats that stay within the space domain, or must retaliatory threats cross domains to be effective?
12. What principles should guide the development of U.S. cross-
domain deterrence policies for space (e.g., are red lines useful,
and, if so, what should they be and how should they be communicated)?
2:15 – 2:30
Break
2:30 – 3:45
Deterring Nuclear Threats in a U.S.–China Crisis or Conflict – Presentation by Michael Chase
Discussion Facilitator – James Scouras
13. Does China share the U.S. goal of isolating the nuclear domain?
14. What U.S. actions in other domains could provoke Chinese nuclear threats or use?
15. What Chinese actions in other domains could provoke U.S. nuclear threats or use?
16. Do Chinese authors predict that their own or U.S. precision conventional, cyber, or space attacks will elicit a nuclear response?
17. Should the United States explicitly retain or relinquish the option to respond to cyber and space attacks with nuclear
weapons, cultivate ambiguity, or be silent on the matter?
3:45 – 5:00
Future Research – Edward Smyth
18. What else does the United States need to understand about China?
19. Are there specific scenarios that should be examined?
20. What has not been considered?
All sessions except the last were conducted as follows. First, a briefing was presented, with
interactive discussion, to provide the foundation required to address cross-domain issues, in
general, and the posed questions in particular. This was followed by a facilitated discussion of
the posed questions, not all of which were discussed due to time limitations. In the last 5–10
minutes, participants provided computer responses to one or more of the posed questions. We
anticipated that participants would gravitate toward answering those questions about which
they were most knowledgeable.
6
Cross-Domain Deterrence in U.S.–China Strategy
Participants
Participants, listed in Table 1, were drawn from academia, military war colleges, federally
funded research and development centers and other think tanks, and industry. As workshop
organizers, we tried to maintain a balance among the various disciplines required for crossdomain analysis.
Table 1. Workshop Participants
Participant
Affiliation
Acton, James
Carnegie Endowment for International Peace
Blank, Stephen
U.S. Army War College
Bishop, Christopher
The Johns Hopkins University Applied Physics Laboratory
Chase, Michael
U.S. Naval War College
Cheng, Dean
Heritage Foundation
DeSimone, Antonio
The Johns Hopkins University Applied Physics Laboratory
Evans, Dennis
The Johns Hopkins University Applied Physics Laboratory
Hopfinger, Patrick
The Johns Hopkins University Applied Physics Laboratory
Ihde, Alexander
The Johns Hopkins University Applied Physics Laboratory
Kauderer, Todd
The Johns Hopkins University Applied Physics Laboratory
Libicki, Martin
The RAND Corporation
Lieber, Keir
Georgetown University
Mahnken, Thomas
The Johns Hopkins University School of Advanced
International Studies
Manzo, Vincent
Center for Strategic and International Studies
Mastro, Oriana
Georgetown University
Melcher, Gregory
The Johns Hopkins University Applied Physics Laboratory
Morgan, Forrest
The RAND Corporation
Nanos, G. Peter
The Johns Hopkins University Applied Physics Laboratory
Scouras, James
The Johns Hopkins University Applied Physics Laboratory
Smyth, Edward
The Johns Hopkins University Applied Physics Laboratory
Stokes, Mark
Project 2049 Institute
7
Workshop Introduction
Approach to this Report
In publishing this report, our purpose is to
provide a communication link between
workshop participants and report readers.
Thus, our approach can be characterized as
minimalist. For each of the major sessions,
we first summarize the oral presentations,
including the discussions they generated.
We next present edited computer responses,
8
excluding only those inputs we judged not
directly relevant to the question and those that
added little insight (e.g., one-word answers).
Finally, we group these responses into major
themes, quoting participants extensively.
Once the draft report was completed,
we sent it to all workshop participants for
review and for clarification or elaboration
of comments they had made during the
workshop.
2
CHINESE PERSPECTIVES
ON DETERRENCE
Our first session focused on Chinese
perspectives on deterrence. There are at least
three distinct reasons to study this topic. The
most direct reason is that, because deterrence
is a strategy that works in the minds of
adversaries to influence decision making,
we must try—to the extent feasible—to
anticipate how our deterrence rhetoric and
actions will play out in the minds of Chinese
leadership and other audiences. Should
that level of understanding prove imperfect,
at a minimum our deterrence policies
should reflect awareness of our limited
understanding of Chinese perspectives. In
either case, being aware of what we know
and what we do not know can, at least in
theory, help to support development of
an effective deterrent by increasing the
credibility of deterrent threats, preventing
the misinterpretation of deterrence signals
sent and received, and averting unintended
escalation. Second, to the extent that
Chinese perspectives on deterrence differ
from those prevalent in the United States,
we might gain a great understanding. Finally,
understanding
Chinese
perspectives—
starting with terminology and associated
definitions—can facilitate international
dialogue on these issues.
Section A summarizes the presentation
on Chinese perspectives on deterrence
provided to the workshop, followed by
Section B on key questions and participant
responses.
A.Presentation Summary:
Chinese Perspectives on
Deterrence
by Thomas
Mahnken
Professor Mahnken gave a presentation on
Chinese views of deterrence, drawn largely
from Chinese military writings. He argued
that Chinese military authors conceive of
deterrence differently than strategic thinkers
in the United States. Specifically, he noted
that Chinese definitions of deterrence place
a greater emphasis on coercion, and some
varieties of deterrence include the use of
force. In addition, he emphasized that
the People’s Liberation Army concepts of
deterrence in each domain differ from those
espoused in the United States, and People’s
Liberation Army concepts on the relationship
between different domains also differ. This
can be seen most starkly in the concept of
“dual deterrence,” which features the use of
nuclear missiles to deter and conventional
missiles to launch strikes.
Definitions of Deterrence
Chinese definitions of deterrence differ from
those of Western thinkers in several ways.
First, Chinese concepts of deterrence are
more active than those of Western strategists,
emphasizing coercion as much as deterrence.
For example, The Science of Military
Strategy (2005) states, “deterrence plays two
9
Chinese Perspectives on Deterrence
basic roles: one is to dissuade the opponent
from doing something through deterrence,
the other is to persuade the opponent what
ought to be done through deterrence, and
both demand the opponent to submit to
the deterrer’s volition.”3 Second, Chinese
military authors view deterrence broadly,
emphasizing the role of concealment,
surprise, and psychological warfare. Third, in
some instances, Chinese authors view the use
of force as part of deterrence. For example,
the authoritative Science of Campaigns
(2006) states, “That is, we use strong military
attack as the backing to create powerful
deterrence, forcing the enemy to give up the
attempt to resist.”4 Indeed, Chinese military
theorists see deterrence and warfighting as
dialectically unified.5
Chinese authors do not write about “crossdomain deterrence” as used in the U.S. policy
community. Chinese military doctrine does,
however, contain discussions of “integrated
strategic deterrence.” As the authors of The
Science of Military Strategy note, different
countries have different means at their
disposal to deter. China, for example, has
nuclear weapons, conventional power, and a
people’s war capability. “By combining these
means of deterrence, an integrated strategic
deterrence is formed, with comprehensive
national power as the basis, conventional
force as the mainstay, nuclear force as
the backup power, and reserve force as
the support.”6
Chinese views of nuclear deterrence
are operationalized in Chinese military
doctrine. Specifically, the People’s Liberation
Army develops operational plans based
on a series of canonical “campaigns.”
More specifically, People’s Liberation
Army
doctrine
includes
“deterrence
campaigns” as well as “warfighting
campaigns.” Deterrence and warfighting
campaigns exist for both nuclear weapons
and
conventional
ballistic
missiles.
Information operations are conceived of
as elements of broader campaigns. Tasks
include protecting campaign information
systems, collecting intelligence, destroying
enemy information systems, and weakening
an enemy’s ability to use information during
war. As yet, however, there appears to be no
independent campaign for the employment
of space forces in Chinese doctrine.
Nuclear Deterrence
Chinese military writings link China’s
nuclear strategy to its status as the object
of superpower nuclear threats. As the
Chinese Military Encyclopedia states, “The
nuclear weapons that China developed
were done under coercion, in order to
break the superpowers’ nuclear monopoly,
to oppose nuclear blackmail, for defense,
and to guard the Chinese people against
the threat of nuclear war.”7 In operational
terms, the Chinese Second Artillery Force
has traditionally thought of and planned
for a number of deterrence and warfighting
Peng Guangqian and Yao Youzhi, The Science of Military Strategy (Beijing: Military Science Publishing
House, 2005); as cited in Dean Cheng, “Chinese Views on Deterrence,” Joint Force Quarterly 60, 2011, 92.
3
Zhang Yuliang, Science of Campaigns (Beijing: National Defense University Press, 2006), 203.
4
Peng Guangqian and Yao Youzhi, The Science of Military Strategy, 171.
5
Ibid, 177–178.
6
Qian Gui and Shen Kehui, “Nuclear Strategy,” Chinese Military Encyclopedia, Song Shilun and Xiao Ke,
eds., Vol 2 of 11 (Beijing: Military Science Publishing House, 1997), 244.
7
10
Cross-Domain Deterrence in U.S.–China Strategy
campaigns: nuclear deterrence, counternuclear deterrence (that is, resisting
intimidation and coercion), and nuclear
counter-attack operations. Since 1994, the
Second Artillery has added a con­ventional
mission, including conventional missile force
deterrence operations, the conventional
missile strike campaign, and support to joint
operations.
The Second Artillery’s Science of
Second Artillery Campaigns (SSAC) outlines
a set of deterrence campaign methods:
(1) exert pressure through public opinion
(information); (2) raise the level of weapons
preparation; (3) demonstrate strength;
(4) create momentum with troops—
feints and simulated and real launches;
(5) conduct launch exercises; (6) conduct
nearby test launches; and (7) reduce the
nuclear threshold. The concept of “reducing
the nuclear threshold” is particularly
important and refers to threatening nuclear
escalation in the face of a conventional
precision-strike campaign by a superior
adversary in the following circumstances:
• “First, when an enemy threatens to
carry out conventional strikes against
our nuclear facilities (or nuclear power
stations),”
• “Second, when an enemy threatens
to carry out strikes against our major
strategic targets related to the safety of
the people, like large-scale water and
electricity stations,”
• “Third, when an enemy threatens to
carry out medium or high strength
conventional strikes against our capital,
important nuclear facilities, and other
political and economic centers,”
• “Fourth, when conventional warfare
continues to escalate and the overall
strategic situation changes from positive
to disadvantageous for us, and when
national safety is seriously threatened,
in order to force the enemy to stop its
invasion and in order to save the country
from danger.”8
Conventional Deterrence
In 1993, the Central Military Commission
assigned the Second Artillery the mission of
“dual deterrence and dual operations,” which
emphasizes the importance of deterrence
and combat roles for both the conventional
and nuclear missile forces. The objective
of conventional missile force deterrence
operations is to influence the enemy’s
decisions by convincing them that China’s
missile force has powerful strike capabilities
and that Beijing has the will to use them
if necessary to prevent the enemy from
challenging China’s interests or to compel
the enemy to accept Beijing’s demands.
Chinese authors write about conventional
missile force deterrence operations of varying
intensity, with associated activities, some of
which include the use of force:
• Low-intensity conventional missile force
deterrence “usually does not have a
very strong confrontational nature.” Its
activities include:
- Using the media to transmit
propaganda about the missile force
and changing the disposition of the
conventional missile force units.
- Continuously improving the missile
force’s survivability, rapid response
capability, ability to penetrate
Second Artillery Corps of the People’s Liberation Army, Science of Second Artillery Campaigns (Beijing:
People’s Liberation Army Press, 2004), 294.
8
11
Chinese Perspectives on Deterrence
missile defense
destructiveness.
systems,
and
• Medium-intensity conventional missile
force deterrence has a “definite confrontational quality.” Its activities include
conducting conventional missile force
exercise launches.
• High-intensity conventional missile
force deterrence has a “very strong
confrontational nature.” It is implemented
through “close proximity or critical
deterrence strikes,” which involve firing
missiles toward an area near an enemy
state or into the waters off of an enemyoccupied island to cause the enemy to feel
an even greater sense of psychological
pressure.9
Space Deterrence
Chinese views of space warfare and space
deterrence have evolved in the last decade
and a half. The entry on “Space Warfare”
in the 1997 edition of the Chinese Military
Encyclopedia viewed military operations in
traditional terms, arguing, “Space warfare
is an extension of warfare on the ground, at
sea, and in the air; and the space force is an
extension of the army, navy and air force.
The emergence of space warfare will have
a certain effect on wars of the future, but it
will not play a decisive role. The primary
factors for deciding victory and failure in
war will remain the nature of the war and
the support or opposition of the people.”10
The 2005 edition of The Science of Military
Strategy portrayed space deterrence in
considerably different terms, arguing that
“it has great effects of shock and awe on the
enemy.” In the view of the authors, the means
of space deterrence are flexible and include
“interference, disruption, and destruction.”
They also note that space deterrence is
restrained by international opinion, space
law, and regulations.11
Information Deterrence
The use of information to deter or compel an
adversary has been a feature of Chinese military
thought for millennia. Indeed, the authors of
The Science of Military Strategy (2005) invoke
Sun Tzu in discussing the topic: “The best
result information deterrence pursues is to
‘subdue the enemy without fighting’ (Sun Tzu)
and strive for winning the victory of war by
confrontation without shedding blood.”12
Chinese authors see information
operations as being launched at the begin­
ning of a conflict and continuing throughout
its course. They view their potential enemies,
including the United States, as information
dependent. They see information operations
as a preemptive method that can be used
to achieve information dominance. They
believe that information operations will
allow China to fight and win an information
campaign, precluding the need for
conventional military action.
Zhao Xijun, Coercive Deterrence Warfare: A Comprehensive Discussion of Missile Deterrence (Beijing:
National Defense University Press, 2005), 171.
9
Yang Zhongcheng, “Space Warfare,” Chinese Military Encyclopedia, Song Shilun and Xiao Ke, eds.,
Vol 3 of 11 (Beijing: Military Science Publishing House, 1997), 602.
10
Peng Guangqian and Yao Youzhi, The Science of Military Strategy (Beijing: Military Science Publishing
House, 2005), 176.
11
Ibid.
12
12
Cross-Domain Deterrence in U.S.–China Strategy
B.Chinese Perspectives on
Deterrence: Key Questions
and Participants’ Responses
This section presents participants’ per­
spectives on the agenda questions associated
with the workshop session titled “Chinese
Perspectives on Deterrence.” The two sources
of information for this section are (1) group
discussions conducted immediately after the
session presentation and (2) computer inputs
provided at the end of the session.
1. How much confidence can
the United States have in
its understanding of Chinese
perspectives on deterrence, based
on existing literature?
Ten workshop participants addressed this
question, with mixed judgments ranging from
“minimal confidence” to “a pretty decent
level of confidence.” Although there was no
consensus on the answer, the median view
hovered in the neighborhood of “limited
confidence.”
One participant made a distinction
between the academic, policy, and military
communities:
• Academics might feel relatively more
confident about U.S. understanding
of deterrence, but I doubt whether the
policy or military communities feel this
way. I certainly believe they should not
feel confident about U.S. insight into
Chinese views.
Several participants focused on the facts
that doctrine may evolve over time and some
of the more important literature is some­what
dated:
• Some of the main published sources are
from about 2000–2006. With some of
these publications now 7–13 years old,
it’s worth keeping in mind that many
things have changed since they were
released, so we should definitely be on
the lookout for new editions of some
of the key sources and/or other new
publications.
• It is hard to know how much a 10-yearold unclassified document, even
if “authoritative” at the time it was
published, accurately reflects the thinking
of current top leadership in the Chinese
government/military.
A number of participants cautioned
against relying too much on doctrinal
literature to predict Chinese actions in
a crisis:
• Do China’s leaders, themselves, really
know when push comes to shove that
they would do more, less, or the same as
they hint they would do?
• People’s Liberation Army doctrine will
not, in the end, determine Central
Military Commission decision making
in the context of a future, hypothetical
contingency.
• Doctrinal literature is just that—doctrinal
literature. It discusses a lot of interesting
things, but it doesn’t necessarily describe
how China would actually act in a real
crisis or conflict.
Final points made were that the Chinese
government is not a monolithic entity and
that we have different levels of insight into
different elements of the bureaucracy:
• While there is a significant body of
literature published by the Academy
of Military Science and the People’s
Liberation Army National Defense
University and Second Artillery Corps,
what do the People’s Liberation Army
General Political Department, the
Central Military Commission General
13
Chinese Perspectives on Deterrence
Office Research Bureau, and the Chinese
Communist Party Central Committee
General Office think about deterrence?
2. Is it an important issue that the
United States and China have
similar, yet different, definitions of
deterrence?
Divergent views were evident from the eight
respondents to this question. Half answered
in the affirmative:
• This is absolutely critical due to the
wide gaps [compared to U.S. thinking]
in Chinese thinking about many of the
concepts that comprise deterrence,
e.g., the role of nuclear weapons and
proportionality, as well as many other
aspects of deterrence.
• This issue is absolutely essential. Different
concepts, different definitions, different
cultures, and different histories shape and
mold understandings and perceptions.
• Yes. The U.S. focus is on capabilities,
while the Chinese focus is on resolve.
Investments by the United States in
capabilities have value for deterrence
only if the Chinese are convinced the
United States has the will to act.
• To the extent that deterrence is based
upon perception, the fact that the United
States and China have differing definitions
of deterrence matters a great deal. To
cite one example, China might launch a
missile in close proximity to U.S. naval
forces as a deterrent signal, whereas the
United States might view such an act as
one of aggression.
The other half of the respondents
acknowledged differences in U.S. and
Chinese definitions of deterrence but
questioned the significance of these
differences:
14
• Yes, but I think the differences are
sometimes exaggerated.
• Not necessarily. It is not impor­
tant where the United States and
China draw the borderline between
“deterrence” and “coercion,” provided
the United States understands Chinese
viewpoints and has an approximately
accurate ability to know what they would
do in a crisis and an ability to predict how
willing they would be to initiate a crisis.
• Too much is made of perceived
differences between U.S. and Chinese
definitions of deterrence. For one
thing, definitions are just definitions.
However, more importantly, I suspect
U.S. and Chinese views are quite similar,
controlling for differences in strategic
position. For example, the Chinese
likely think of nuclear use in ways far
more similar to the United States than
one might conclude from public debates
about China’s “no first use” policy.
• Yes, but I’m not certain of the
ramifications. The Chinese view of
“correlation of forces” as ultimately
determinative is not the same as the
Western notion of matching ends
and means or stopping at less than
ultimate victory in the pursuit of longrun stability.
3. What misperceptions exist in China
and the United States about each
other’s deterrence strategy?
The motivation for this question is the concern
that misperceptions about each other’s
deterrence strategy could inadvertently
lead to conflict and/or conflict escalation.
Misperceptions identified by participants
focused on identity, motivation, and the
potential for escalation:
Cross-Domain Deterrence in U.S.–China Strategy
• One critical issue is that, in a major conflict,
both the United States and China could
perceive themselves as the “defender,”
with the consequence that escalation
would become more likely and harder to
control. This is reflected in the language
used by each side to describe capabilities
to hinder U.S. freedom of access; the
United States calls them “anti-access/areadenial capabilities,” while China calls
them “counter-intervention capabilities.”
More generally, China views its build-up
as defensive and the goal of reunifying
Taiwan as correcting an historical
injustice. The United States worries the
build-up is offensively oriented and would
view an attempt to seize Taiwan by force
as offensive. This is a recipe for escalation.
• Some Chinese general officers still think
that the goal of the United States is to destroy
them. This was shockingly apparent when
a group of Chinese generals and colonels
from their nuclear rocket forces met with
Department of Defense personnel in 2008
or 2009.
• I think U.S. civilian and military planners
misperceive the degree to which AirSea
Battle will trigger Chinese escalation.
I think Chinese leaders misperceive
the willingness of the United States to
engage in military conflict in East Asia
over issues that China perceives as
non-vital/non-strategic to the United
States. Paradoxically, the Chinese also
probably misperceive U.S. willingness to
revise the status-quo (i.e., they probably
believe that we have more aggressive/
revisionist designs than we have).
• Many misperceptions exist; e.g., under
what conditions would each side go
nuclear, how nuclear weapons deter
reprisals for conventional or information
warfare strikes, what constitutes a first
strike, etc.
• Misperceptions exist regarding issues of
resolve and controllability of escalation
and crisis management. Do crises develop
momentums of their own? It is unclear
whether the Chinese fear loss of crisis
control as much as the United States does.
One participant noted the difficulty
faced by China due to multiple voices in the
United States:
• The Chinese have a hard time separating
official U.S. policy and what they may
infer from statements by U.S. media,
Congressmen, lower-level U.S. officials,
or Defense Science Board reports.
4. What do the Chinese think of the
U.S. discussion of “cross-domain
deterrence”?
No respondent identified any Chinese
mention of the U.S. discussion of crossdomain deterrence:
• I am not aware of any Chinese reactions
to U.S. discussion of cross-domain
deterrence.
• I haven’t read anything yet in Chinese
literature about this.
Some explained China’s thinking without
reference to the U.S. discussion:
• China considers strategic deterrence to
involve all elements of national power
in a seemingly progressive pattern. How
does this contrast with the U.S. process?
China’s view incorporates all elements of
national power and, unlike ours, elides
the difference between peace and war.
• Chinese writings suggest that they
believe in cross-domain deterrence.
I would suggest the focus is not
“cross-domain,” but that it is holistic—
deter­rence
across
domains,
but
deterrence isn’t stovepiped to begin with.
15
Chinese Perspectives on Deterrence
• My impression is that it isn’t thought of as
a separate concept. From the perspective
of a weaker power, asymmetry is the goal.
The ability to leverage capabilities in all
domains to inspire the desired effect is
a given. Cross-domain deterrence may
be considered more robust only because
employing multiple domains could
be considered a force multiplier,
i.e., more effective militarily, and
therefore, the threat of it should enhance
deterrence.
One respondent took note of the
incoherence of thought in the United States
about cross-domain deterrence and, by
implication, the difficulty China would have
of making sense of it:
• I wonder if anyone could coherently
identify or summarize the U.S. discussion
of cross-domain deterrence.
5. China considers strategic deter­
rence to involve all elements of
national power in a seemingly
progressive pattern. How does this
contrast with the U.S. process?
Most respondents’ comments suggested
that Chinese conceptions about deterrence
are both broader and more integrated than
those of the United States. By contrast, most
thinking in the United States about deterrence
focuses on the nuclear domain:
• It would be great if the United States had
a coordinated, whole-of-government
approach to strategic deterrence, but
it doesn’t. The United States has in
recent years assigned “deterrence” to
combatant commands, where it is dealt
with variably.
• U.S. leaders tend to think of deterrence in
a more compartmented fashion, focusing
mainly on nuclear. This wasn’t always
16
true, of course. Throughout most of the
Cold War, the United States used threats
of nuclear retribution to deter a Soviet
conventional attack on Western Europe.
In the post-Cold War era, however, U.S.
leaders have largely assumed that its
conventional forces would prevail in any
war it chose to enter. Consequently, they
have given much less thought to deterring
conventional attacks. They have also
attempted to segregate nuclear threats
and marginalize them in international
discourse, believing such actions reduce
risks of nuclear war and pressures for
nuclear proliferation. Chinese and
Russian leaders, however, see such moves
as thinly veiled attempts to secure U.S.
advantages in conventional warfighting.
Consequently, they do not accept this
strict segregation. Just as when the United
States believed itself conventionally
vulnerable, China and Russia have turned
increasingly to integrating their nuclear
and conventional deterrence concepts.
• Too often, in the United States, deterrence
is meant to be nuclear.
• For China, national power is not
progressive. It is much more holistic. It is
not politics to diplomacy to economy to
military. It is, as the Chinese say, a matter
of “comprehensive national power.”
• The United States is more likely to
separate the various notions of deterrence
and limit “strategic deterrence” to
nuclear war. However, it is unclear what
difference this distinction makes.
Two respondents did not see a difference
or thought suggested differences may be
exaggerated:
• Both the United States and China regard
deterrence as involving all aspects of
national power in a somewhat progressive
manner.
Cross-Domain Deterrence in U.S.–China Strategy
• The United States casts China in some
sort of idealized light, believing that all
its strategies and policies are integrated
while ours are stovepiped. People
commented on U.S. resolve, but Chinese
resolve (China’s willingness to act and
disregard the effectiveness of U.S.
and world responses) should also be
considered.
6. People’s Liberation Army doctrine
consists of both “deterrence
campaigns” as well as “warfighting
campaigns.” Is this in contrast to
U.S. doctrine?
Respondents identified the following
differences between U.S. and Chinese
doctrine:
• The United States does not really think
about campaigns quite the way the
Chinese do; in other words, it doesn’t
have a “joint blockade campaign” or
“anti-air raid campaign.” So there is
a contrast, but it is broader than the
distinction between deterrence actions
and warfighting campaigns.
• China’s concept of active deterrence
is similar to Schelling’s concept of
suggestive escalation and threats that
leave something to chance—using force,
perhaps against a tactical target, primarily
to influence the adversary’s perceptions
(e.g., to deter escalation or increased
involvement in the conflict). This concept
has soaked into U.S. thinking about
limited nuclear strikes.
• Chinese doctrine is more blurred,
perhaps because of less concern about
the inability to sustain crisis management.
• The United States thinks about deterrence
as a “Phase 0” problem. It also does not
integrate deterrence and warfighting to
the extent that China does.
One respondent emphasized similarities
over differences:
• I think observers tend to exaggerate
the differences here. The United States
tends to talk about military planning
and force requirements for the purpose
of deterrence and, if deterrence fails,
for mitigation (which essentially means
warfighting). The fact of the matter is
that warfighting military capabilities
are arguably the key building blocks of
deterrence, and this is recognized by
both U.S. and Chinese leaders.
17
3
DETERRING CHINESE
CYBER ATTACKS
Deterring Chinese cyber attacks is, arguably,
the most daunting of the cross-domain
challenges facing the United States. This is
due to the confluence of several phenomena,
in addition to the oft-cited asymmetric U.S.
dependence on cyber for both military and
civilian applications and the difficulties with
attribution of cyber attacks. First, cyber is the
newest domain of warfare and, as such, lacks
the maturity of thinking that has been applied
to many other domains of warfare, particularly
the nuclear and conventional domains. In
addition, cyber intrusions (as opposed to
attacks that cause tangible damage to systems
or people) have become commonplace,
creating a precedent that makes establishing
contrary norms of behavior far more difficult
than, for example, in the space domain,
which has, since Sputnik, enjoyed the norm
that peacetime attacks on satellites are
neither expected nor acceptable behavior.
Finally, there is little experience with cyber
attacks that would establish a solid basis for
expectations regarding retaliation.
Section A summarizes the presentation
on deterring Chinese cyber attacks provided
at the workshop, followed by Section B on
key questions and participant responses.
A.Presentation Summary:
Responding to Chinese
Information Warfare
by
Stephen Blank
Dr. Blank’s presentation is based on a draft
paper supplied in advance to workshop
participants.13 He first discussed Chinese
perspectives
regarding
the
nuclear,
conventional, cyber, and space domains,
emphasizing the inherently cross-domain
nature of Chinese deterrence strategy
and military operations planning. After a
brief discussion of new Chinese doctrinal
concepts, he concluded with suggestions for
counteracting approaches that the United
States can take to enhance deterrence and
prevail in war should deterrence fail.
Chinese Domain Perspectives
Nuclear. China uses nuclear weapons as
a shield for its information warfare and
other preemptive offensive operations (or,
as China describes such actions, “active
defense”) by deterring retaliatory attacks
in all domains—nuclear, conventional,
and even information warfare. Just as every
element of national power is seen by China
Stephen Blank, Responding to Chinese Information Warfare, U.S. Army War College, 2013, unpublished.
13
19
Deterring Chinese Cyber Attacks
as contributing to deterrence, China’s
nuclear deterrence applies to all forms of
enemy retaliatory action. For example,
consider a scenario in which China executes
a cyber attack that knocked out U.S. satellites.
China might well brandish nuclear weapons
to deter conventional or information warfare
reprisals.
Thus, nuclear deterrence is inherently
a cross-domain form of deterrence. This
assessment is based on an analysis of
literature by American, Indian, and Chinese
authors, as well as consideration of points of
similarity with Russia, where conventional
attacks on, for example, command, control,
communications, computers, intelligence,
surveillance, and reconnaissance (C4ISR)
would clearly provoke a nuclear response.14
An important implication of China’s
strategy of using nuclear weapons to deter
attacks from other domains is that Chinese
proclamations about no first use cannot be
relied on.
Conventional. Regarding the conven­
tional domain, it is important to recognize
that informational capabilities now permeate
all Chinese military operations in land, air,
sea, and space. They are now approaching
the point where conventional operations
are inherently cross-domain. Just as it has
made the United States vulnerable, growing
reliance on informational capabilities
to support conventional operations is now
also making China vulnerable to information
warfare.
Information Warfare. Chinese views
on information warfare differ in important
ways from those of the United States. Similar
to nuclear weapons, and unlike the United
States, China views information warfare as
providing a form of deterrence against all
manner of enemy reprisals—informational,
conventional, and even nuclear. Computer
network attack is the spearpoint of this
deterrent.
While both the United States and China
use information warfare in Phase 0, the
United States sees peace as one thing and
war as something completely antithetical,
in contrast to the Chinese perspective of a
much more seamless connection between
peace and war. Thus, the Chinese view
Phase 0 information warfare as part of
intelligence preparation of the battlefield
and as including perception management
on a grand scale. It is also very possible
that China sees information warfare as an
assassin’s mace—a weapon that can strike
suddenly and decisively to overcome an
enemy’s strength.
Space. Space is of increasing importance
to China, as evidenced by anti-satellite
activity; civilian space efforts; and support
of conventional military capabilities to wage
information warfare, launch strikes to target
submarines, for example, and deter enemy
strikes and counter-strikes against Chinese
space and other systems. This reinforces the
conclusion that Chinese military operations,
like those of the United States, are inherently
cross-domain and becoming more so
over time.
Cyber, space, air, and naval capabilities
are
increasingly
organically
linked
and interdependent; thus, a successful
information warfare campaign could
seriously degrade military capabilities. Both
sides are increasingly reliant on cross-domain
operations and vulnerable to information
warfare campaigns intended to impede and
degrade the other side’s operations.
One participant cautioned that many experts believe China is still focused on nuclear weapons as
deterring only, or primarily, nuclear use.
14
20
Cross-Domain Deterrence in U.S.–China Strategy
New Doctrinal Concepts
There are new Chinese doctrinal concepts
developing, which reinforce the inherently
cross-domain nature of military operations.
Integrated electronic warfare pulls together
all operations utilizing information warfare
against enemy targets. Closely aligned with
the People’s Liberation Army’s doctrine for
fighting local warfare for limited objectives
under informatisized conditions, this doctrine
calls for a fully networked architecture capable
of coordinating military operations in all
domains and across the electronic spectrum,
including electronic warfare attacks on enemy
C4ISR networks and computer network attacks
on information networks.
A related Chinese concept, known
as “systems sabotage warfare,” involves
degrading the enemy’s C4ISR to such a
degree that it will lack the capability or the
will to resist Chinese operations. This concept
is more evidence of inherently integrated
activities across the electromagnetic
spectrum and all domains.
How Should the United States
Respond?
Given that the Chinese have made it clear
that they seek to prevail in localized warfare
for limited objectives under informatisized
conditions, how should the United States
respond? First, the United States must make
clear to China that any such conflict involving
U.S. interests and allies, or the United States
itself, would be prolonged. China fears internal
security tensions (social and ethnic unrest,
economic slowdown, etc.) more than external
threats. A prolonged war would aggravate
these domestic tensions. For example, actions
such as an oil blockade in the Malacca Strait,
while not necessarily the answer, could
prolong a conflict and aggravate domestic
economic and political tensions in China.
Second, the United States must also
invest in robust and redundant U.S. and
allied conventional, space, missile, and
cyber defenses. Such a capability would
make it difficult for China to seek a quick
victory. In addition, it would also help the
United States to prevail in a prolonged war
and thus would also enhance deterrence
of China. For example, China may well be
miscalculating the impact of cyber strikes on
U.S. or allied targets. It is unclear whether
Chinese threats of cyber strikes alone, even
if they are “overwhelming,” would deter
adversaries. However, in any event, robust
cyber defenses would send a signal that not
only can the United States prolong the war,
but Chinese cyber threats would not deter
U.S. and allied responses.
Taking this one step further, beyond
working with allies to develop an Asian–
Pacific intelligence, surveillance, and
reconnaissance (ISR) capability, the United
States could also work to develop an
information strike capability. While this
might provoke a reaction from China,
it should reinforce the perception that China
will not be able to achieve a quick victory.
Third, the United States needs to be able
to conduct offensive conventional strikes
on Chinese ISR capabilities. However, there
is an inherent danger in this approach.
Although offensive conventional strikes may
be successful in wartime in crippling Chinese
military capabilities, such strikes could
provoke China into escalation that would
lead to nuclear war.
Strikes of C4ISR targets, especially,
could well meet the criteria to abandon
China’s policy of no first use. It is clear from
Russian doctrine that striking such targets
would meet the Russian criteria. Given
the contemporary strategic environment,
systematic large-scale attacks on C4ISR
could lead to an escalation at the nuclear
21
Deterring Chinese Cyber Attacks
level. Therefore, there is an inherent crossdomain risk in those attacks. This could be
true for both sides, and the Chinese may
not be fully aware that they are playing
with fire.
Fourth, the United States needs to
look to diplomacy to complement military
actions and responses. Together with allies,
the United States should develop/build
a network of relationships and capabilities
to enhance ISR and deter Chinese probes.
The United States also needs to include
cyber issues (as well as cross-domain and
nuclear) as part of military discussions
with China, just as nuclear issues have
been discussed with Russia for the past
55 years. China needs to understand
the United States better, just as the
United States needs to understand
China better.
Finally, the United States should agree
with Russia that there will be no subsequent
strategic arms reduction talks without
including China. Having the Chinese
involved in such a discussion would (1)
force discussions of strategic contingencies;
(2) create a process to develop mutual
understanding over time; (3) stop China
from being a free-rider on the superpowers;
(4) reduce opacity in Chinese nuclear
capabilities and plans; (5) politically, drive
a wedge between Moscow and Beijing; and
(6) bring everyone into a more secure and
regularized relationship.
B.Deterring Chinese Cyber
Attacks: Key Questions and
Participants’ Responses
This
section
presents
participants’
perspectives on the agenda questions
associated with the workshop session titled
“Deterring Chinese Cyber Attacks.” The two
sources of information for this section are (1)
group discussions conducted immediately
22
after the presentation and (2) computer inputs
provided at the end of the session.
7. Can the United States effectively
dissuade cyber attacks without
demonstrating willingness to retaliate
(i.e., are protection strategies and
establishing
behavioral
norms
sufficient for preventing cyber
attacks)?
Two respondents based their answers on a
fundamental tenet of deterrence—if there is
nothing to be gained by attacking, there is no
motive to attack and thus no need to threaten
retaliation to deter an attack. However, these
respondents appear to differ on the likelihood
of achieving defenses that can achieve the
necessary level of protection. The first sets
a very high standard—defenses must be
“virtually impenetrable”—and implies that
such a standard is impossible to meet:
• No. Unless U.S. cyber defenses are
virtually impenetrable, to deter a cyber
attack, the perpetrator must be at risk of
losing something that it already has.
The second respondent set a more
ambiguous standard—defenses must be “good
enough”—and implied that such defenses
might be feasible:
• If your defense is good enough, it doesn’t
matter that someone else is attacking. In
the long run, sufficient levels of frustration
on the attacker’s part may dissuade
them from investing in their search for
vulnerabilities in your systems.
Several respondents emphasized that the
answer to this question turns on the meaning of
“attacks.” Attacks that cause damage to either
military capabilities or civil infrastructures
must be distinguished from cyber espionage
intended to exfiltrate data and other forms
of cyber intrusion. These respondents agreed
Cross-Domain Deterrence in U.S.–China Strategy
that retaliatory threats would be necessary to
deter cyber attacks but might be inappropriate
or inadequate against cyber intrusions.
• It is important to distinguish between
cyber attacks aimed at inflicting actual
damage and cyber espionage. Espionage
is to be expected and probably should
not provoke retaliation against a foreign
country.
How could you deter a cyber attack
without demonstrating a capability and
the will to retaliate? Computer network
defense and norms alone seem unlikely
to be enough or at least not without some
threat of retaliation in kind or through
some other means.
How could cyber espionage be deterred?
It probably cannot be deterred with
defenses, and threat of retaliation in kind
will not help if the other party assumes
similar activity is already underway.
Norms might have a role but would require
agreement between the parties about the
norms that would be appropriate. Maybe
other means—economic or diplomatic
consequences,
most
likely—could
influence China’s decision calculus and
deter at least some types of cyber espionage
(such as theft of intellectual property and
confidential business information from
U.S. companies).
• This is a really tough question, especially
given the ambiguity of what “cyber
attacks” might actually entail. I don’t think
a “demonstrated” willingness to retaliate
is necessary for effectively dissuading
cyber attacks on U.S. military systems and
capabilities that would effectively cost
U.S. lives or undermine military missions
during a conflict. They should expect that
retaliation. However, if the goal is deterring
the kinds of cyber attacks that have already
occurred (i.e., largely of the espionage
type), the United States would need to
dramatically ramp-up its own offensive
cyber operations against Chinese targets
and do so in a way that clearly conveys
that such actions are explicitly retaliatory
(and even escalatory).
• Yes, but only if the United States can
demonstrate a consistent ability to
promptly detect and defeat attacks and
close the gaps through which they came
without suffering degraded operational
capabilities. Opponents can launch
cyber attacks with very low costs to
themselves. Therefore, the main costbenefit calculation opponents have to
consider is whether the degradation
in U.S. operational capability (or the
value of cost imposed on the United
States) would be great enough to offset
the attacker’s cost in terms of the loss
in intelligence exploitation it had been
enjoying in the targeted system up to
the point of the attack. If the United
States can make defenses and forensic
capabilities sufficiently responsive and
reliable to deny an opponent’s benefits
and promptly close off access, then
calculating state adversaries will not
attack. The United States will still be
plagued by hacker groups and others just
interested in causing trouble, but serious
threats will be deterred.
One respondent emphasized that cyber
attacks are most likely to occur not in
isolation, but as part of a larger conflict, and
that the effectiveness of deterrence should be
considered in that broader context:
• A cyber attack is likely to occur as part of a
broader attack. Demonstrations of resolve
and willingness to retaliate to an attack
more generally should dissuade cyber
attack as a subordinate element of such
an attack.
23
Deterring Chinese Cyber Attacks
One respondent thought that only
demonstrated protection and recovery
capabilities would be effective, while another
thought that only retaliation would be
effective:
• It depends what you believe about how
China thinks about using cyber. I don’t
think the Chinese are thinking of it to deter,
i.e., to threaten use of cyber to harm the
opponent to convince the opponent not
to take some action. I think China thinks
about using cyber to reduce U.S. military
effectiveness—to degrade, delay, deny,
etc. U.S. use of force. If this is true, the
best way to deter Chinese cyber attacks is
to demonstrate (1) the United States can
operate with limited C4ISR; (2) if it can’t,
it will wait until it is back online and then
act; and (3) the United States can protect
itself. I think behavioral norms won’t help,
and U.S. ability to retaliate with cyber
won’t do it either.
• Obviously not. Current thinking concludes
that the United States is not capable of
protecting its society against a severe
information attack, and such robustness
is years in the future, if it is possible at
all. Because of the challenges related to
attribution, U.S. unwillingness to agree to
international norms, and the newness of
the technology, international behavioral
norms have not arisen and are not likely
to be codified in the near future. In the
international community, it is expected
that the United States retaliates in kind
to attacks—disruption for disruption, loss
of life for loss of life. The expectation that
the United States would conduct such an
attack in retaliation serves as deterrence.
8. Are there effective retaliatory
threats that stay within the cyber
domain, or must retaliatory threats
cross domains to be effective?
24
Most respondents thought that limiting
retaliatory threats to the cyber domain would
not be effective in deterring cyber attacks,
largely due to asymmetric U.S. dependency
on information systems. One respondent
emphasized that, with greater understanding
of Chinese perceptions, ambiguity regarding
China’s potential response to a cyber attack
could work to the U.S. advantage:
• Because the United States is more
reliant on cyber capabilities than China,
it seems unlikely that, in general, U.S.
responses could be effective deterrents if
they were limited to the cyber domain.
That said, the United States is not—and
in practice almost certainly will not—
be specific about how it will respond
to cyber attacks. For this reason, the
real challenges are (1) understanding
how China thinks the United States
would respond to a cyber attack and (2)
understanding how to shape Chinese
perceptions in a direction favorable to
the United States.
As with the previous question, some
respondents distinguished cyber espionage
and cyber attacks:
• To deter cyber espionage, the United
States could improve defenses or
reduce the value of the information that
is collected by feeding the intruders
disinformation, but I think you probably
have to go across domains—into the
economic and diplomatic domains, in
particular—to have much of an influence
on the other side’s decision calculus. To
deter cyber attacks, threat of retaliation
in kind might be perfectly sufficient, at
least for some types of attacks.
Again, similar to the previous question,
several respondents emphasized that cyber
attacks are likely to occur in the context of
a broader conflict, with the implication that
Cross-Domain Deterrence in U.S.–China Strategy
responses need not be constrained to the
cyber domain:
• Because cyber attacks are unlikely
to occur on their own, retaliation
is likely to occur in the context of a larger
conflict.
• Because an attack is unlikely to be onedimensional, retaliatory effects need not
be one-dimensional as well.
• There could be proportional or equivalent
cyber retaliatory strikes in peacetime, but
during an actual war, the United States
may well have to employ cross-domain
retaliation because, on both sides, all
military operations either are crossdomain or are increasingly becoming
cross-domain.
Respondents raised a variety of issues
with limiting responses to the cyber domain.
Among them were the U.S. offensive cyber
capability and the level of confidence the
United States can have in it:
• I do not think retaliatory threats “must”
cross domains to be effective, but the
question of whether there are effective
retaliatory threats that stay within
the cyber domain turns on technical
questions about U.S. offensive cyber
capabilities and the degree of damage
the United States can inflict on Chinese
systems and capabilities.
• In principle, cyber retaliation could deter
a cyber attack, but the United States
would need to be able to inflict serious
damage on a perpetrator through cyber
retaliation. In practice, it is uncertain that
the United States can be confident of
such a retaliatory cyber capability, so it
needs to consider non-cyber approaches
(to include economic penalties against
a nation state that institutes a cyber
attack).
One respondent noted the relationship
between the damage that would be inflicted
by a retaliatory threat and the credibility of
that threat:
•A same-domain retaliatory threat
produces weaker effects (or at least no
stronger effects), but it is more credible
than a threat that would have the United
States respond with kinetic weapons
to a non-kinetic attack. It comes down
to the question of whether adversaries
think the United States wants to go
to war over an attack that does not
kill anyone.
Several respondents focused on the issue
of proportionality:
• While it is likely that U.S. retaliation for a
cyber attack would be cyber related, this
is not guaranteed. However, the impact
of the retaliation should be similar to
the kind of effects suffered by the United
States. Examples of acceptable retaliation
to a cyber attack that caused a blackout
in a major U.S. city that lasted 24–48
hours include a U.S. attack that achieved
similar results conducted through agents/
proxies or a cruise missile against a power
distribution station.
9. What principles should guide the
development of U.S. cross-domain
deterrence policies for cyber (e.g.,
are red lines useful, and, if so, what
should they be and how should
they be communicated)?
Respondents discussed a wide variety
of
principles
involving
red
lines,
proportionality,
ambiguity,
attribution,
targets, and cooperation:
• Red lines could be credible if they are
communicated both openly and privately
and then upheld by action in response to
25
Deterring Chinese Cyber Attacks
active probes. The United States would
have to make clear which targets or kinds
of attacks are “off limits.” It should also
make clear that it will retaliate against
attacks.
• Maybe the United States could make a
distinction between tactical targets that
support military operations (the power
to an air defense radar) and strategic
or critical infrastructure targets (the
power to a city) and persuade others to
agree that the former is a legitimate
military target while the latter should be
off limits.
• The United States must make it clear that
a cyber attack aimed at inflicting serious
damage (e.g., crashing the electrical grid
or the stock market) will be treated the
same as a conventional attack using a
cruise missile, especially if such an attack
succeeds. By contrast, cyber espionage
needs to be treated as consistent with
expected international behavior, and
the primary response should probably
be defensive. Of course, the preceding
discussion assumes the United States
knows who took down its electrical grid
through a cyber attack. Unless the attack
occurs in conjunction with some sort of
broader war, it may be very difficult to
know for sure who conducted the attack.
• The United States should first develop a
concept of how cyber warfare is likely
to occur, in what form, and in what
relationship to larger conflict before it
starts talking about red lines. That being
said, deterrence is often enhanced
26
by uncertainty (“the threat that leaves
something to chance”).
• Say nothing explicit. The United States
does not have the attribution capability
to guarantee a response, and it does not
have a history of response to convince
others that it failed to respond due to a
lack of understanding, rather than a lack
in motivation to retaliate.
• An eye-for-an-eye, a tooth-for-a-tooth
would probably work as well for
developing cross-domain deterrence
policies as anything else. That is to say, this
principle would be both more credible and
more effective than attempting to convey
a more escalatory threat. For example, if
the United States wants to deter Chinese
cyber attacks that undermine real U.S.
military effectiveness, cost U.S. lives, or
destroy critical infrastructure, threatening
cross-domain strikes against Chinese
military targets would seem reasonable,
credible, and probably the most effective
strategy for deterring those attacks in the
first place. Issuing those threats to deter
continued espionage would not be an
effective strategy.
• Actions such as deepening and broadening
cooperation with Taiwan in the cyber
domain—a bilateral “cyber coalition”—
could impose a cost on Chinese cyber
operations. Cyber cooperation includes
intelligence sharing, a combined
computer emergency response center,
synchronization of information and
communication technology policy, etc.
4
DETERRING CHINESE ATTACKS
ON U.S. SPACE CAPABILITIES
Deterring Chinese attacks on U.S. space
capabilities has become a more urgent
concern with China’s recent successful antisatellite test. The United States is highly
dependent on space assets for the effective
functioning of its military, including both
conventional and nuclear capabilities.
Although China is also becoming more
dependent on space, the large asymmetry
in space dependency endures. At the same
time, little attention has been paid to making
satellites less vulnerable, increasing the
robustness of constellations of satellites and
maintaining terrestrial alternatives to space
capabilities. Moreover, since Sputnik, while
the precedent of unfettered overflight of
national territory by satellites has become
ever more firmly entrenched in practice, as
a matter of policy, China does not accept
this norm. Thus, in the search for strategies
to protect U.S. space assets, cross-domain
deterrence looms large, albeit highly
problematic.
Section A summarizes the presentation
on deterring Chinese space attacks provided
to the workshop, followed by Section B on
key questions and participant responses.
A.Presentation Summary:
Deterring Chinese Space
Attacks
by
Forrest Morgan
Dr. Morgan first conceptualized space
deterrence as a stability issue, then provided
relevant principles of deterrence, and applied
them to the issue of deterring Chinese threats
to U.S. space capabilities. He concluded
with recommendations on developing a
national space policy, making threats of
retaliation more credible, and enhancing the
ability to reduce benefits of an attack to an
adversary.
Conceptualizing Space Deterrence
as a Stability Issue
Deterring attacks on space assets is to some
degree a first-strike stability issue. In the
late Cold War, the United States thought of
the first-strike stability problem in a game
theoretic framework: if neither side can
manage escalation once nuclear war breaks
out, both sides are motivated to strike first
and do as much damage limitation as early
as they can. While space is not exactly that
world, there will be relatively few and weak
firebreaks against rapid escalation if conflict
breaks out in space.
Deterrence in space and in the terrestrial
domain are interdependent. An earlier RAND
study concluded that the United States can
deter attacks in space in peacetime when
it does not need to, and it cannot during
wartime when it does, so deterrence has no
value. However, the United States should
not think exclusively in terms of space
deterrence. The real issue involves the critical
threshold between peace and war: what role
does deterring others from attacking the
United States in space play in raising that
threshold to a higher level, so we do not go
from peace to war? That is ultimately a crossdomain issue.
27
Deterring Chinese Attacks on U.S. Space Capabilities
To illustrate, if China is at the brink of
war with the United States and it thinks it
can succeed in the terrestrial environment
by going after the United States in space,
and the United States cannot deter China
from that course of action, China is more
apt to cross that threshold. However, if the
United States can structure the international
environment and U.S. space architectures
such that when China looks at the costbenefit tradeoffs, it decides that attacking the
United States in space is a higher threshold
than it wants to cross, then when it looks
at attacking in the terrestrial environment,
it will realize it would have to fight a
U.S. conventional military force with full
transformational capabilities fully supported
by space capabilities, the threshold of that
decision is raised as well. So deterring space
attacks is a critical cross-domain challenge
for general deterrence.
Fundamentals of Deterrence
Deterrence is about manipulating a potential
adversary’s decision calculus about whether
to launch an attack or not. An effective
deterrent will persuade an adversary that
the probable costs of an attack are likely to
exceed the probable benefits. To create this
perception, the United States can either raise
expectations of cost via threats of punishment,
lower expectations of benefits via defenses,
or both. Deterrence and defense, done right,
are mutually reinforcing.
If deterrence involves only military
threats of punishment, rather than including
protection as well, the decision space
narrows down to mainly nuclear deterrence
because the cost of executing those
threats is so high it will give the adversary
dramatic pause.
The history of conventional deterrence
is a very different story. It is very difficult
to deter an adversary from attacking by
28
using threats of conventional punishment.
Look at Slobodan Milošević in Serbia. Look
at Saddam Hussein in Iraq. Deterrence
failed in these cases because the adversary
thought he could withstand conventional
punishment long enough to achieve his
military or political objectives or get
someone to intercede on his behalf. In the
conventional world, in the broader scope of
history, as opposed to just the nuclear age,
deterrence has been mainly about defense.
Applying Deterrence Principles
to Chinese Attack on U.S. Space
Capabilities
Deterring attacks on space assets is more
like conventional deterrence than nuclear
deterrence. It is really hard to deter an
adversary by threatening to impose costs.
Adversaries
will
attack
selected
space capabilities in selected ways to
create operational effects (interdict U.S.
force enhancement capabilities for U.S.
transformational warfighting forces in the
terrestrial environment). For example,
adversaries could attack reconnaissance to
make it hard to target mobile forces; strike
communications to affect U.S. networks;
interdict GPS signals to affect U.S. precision
guided weapons, etc. The key question
the adversary will ask is whether attacking
assets in space makes my terrestrial military
operations more effective.
During the early portion of the Cold War,
the United States had a tacit understanding
with the Soviets not to interfere with each
other’s space assets, except on the margins.
However, both sides then used space assets
mainly to support their national strategic
(i.e., nuclear) missions. After the sobering
effects of the Cuban Missile Crisis, the United
States realized that stability was better than
fighting a nuclear war; therefore, it did not
want to do anything to its space assets that
Cross-Domain Deterrence in U.S.–China Strategy
created the perception that it might be about
to initiate a nuclear war.
The situation has changed over the
years, though. In the last several decades,
the United States has progressively used
space assets to support U.S. conventional
terrestrial military forces and do so in a way
that has made U.S. forces very intimidating to
potential adversaries. As a result, the People’s
Liberation Army might find attacking U.S.
space assets an attractive option in efforts
to “level the playing field” by reducing U.S.
warfighting capabilities.
It is not just that China is developing
a counter-space capability to defeat the
United States in war, demonstrating the
ability to attack in space contributes to
China’s ability to deter U.S. intervention in
the western Pacific. China is developing and
demonstrating counter-space capabilities
to let the United States know what it is up
against if it comes to a fight—that it is going
to take some losses in U.S. space assets. That
will raise U.S. costs in terms of casualties
and losses; therefore, the United States had
better take that into consideration in its
weighing of costs, benefits, and stakes in
deciding whether or not to intervene in, for
example, a Taiwan crisis.
How Does the United States
Prevent China from Attacking U.S.
Space Capabilities?
Both threats of punishment and efforts
to demonstrate U.S. ability to deny
military success have problems. Threats of
punishment suffer from defects in potency
and credibility. Efforts to deny China’s success
are hindered by U.S. system vulnerability.
Both are hampered by U.S. disproportionate
dependence on space.
Punishment. It is hard to make threats
of punishment sufficiently potent. Tit-for-tat
probably would not work, given greater U.S.
dependence on space assets. For example,
the threat of retaliating against Chinese
reconnaissance or communications satellites
if China attacks U.S. space assets could be
perfectly acceptable to China. Although
Chinese forces are becoming increasingly
dependent on space capabilities, even if you
project decades into the future, it is unlikely
that China will become as dependent on
space capabilities as the United States, at
least for conflicts in the western Pacific,
where China can rely on terrestrial assets
for the same purposes for which the United
States uses space.
What about going after anti-satellite
launchers (with stealth fighters or bombers)
if China attacks U.S. satellites? (Some even
propose nuclear reprisals, but that is not
going to happen.) The problem is that U.S.
chances of catching and knocking out mobile
anti-satellite launchers are not high, even
if it can fly in unopposed. Also, China will
likely be launching from western China. Will
U.S. political leaders be willing to escalate
a localized conflict on the Chinese coast to
fight its way in and strike targets deep into
Chinese territory and kill Chinese citizens
because China attacked an unmanned piece
of machinery? For deterrence, can you make
these threats credible?
What about threatening the fixed
infrastructure that enables tracking and
targeting of space systems? Reversible- effects
attacks may be conducted using assets that
are close to the area of combat. However, the
United States will likely have higher priority
targets than, for example, a communications
satellite jammer. Attacks on space assets
in geosynchronous orbit are likely to be
supported by redundant infrastructure spread
out across China. Much of the infrastructure
for tracking and command and control (C2)
is in Chinese cities, including Beijing, which
raises proportionality concerns.
29
Deterring Chinese Attacks on U.S. Space Capabilities
When the war escalates to a high
level, the threat of punishment becomes
largely irrelevant.
Denial. Activities to deny the military
effectiveness of attacks fall into the
categories of passive and active defenses.
Passive defenses are limited and expensive.
For example, while the most critical U.S.
communications—nuclear C2—are jam
resistant, the vast majority are unprotected
communications. The trend over the last
couple of decades has been to contract out
on civilian carriers for communications, but
most communications satellite providers do
not want to pay for passive defenses.
Active defenses (intercept satellites,
escort satellites) have serious challenges in
terms of physics, expenses, and capability.
Escorting satellites have operational
issues—how does an escort satellite actually
protect another satellite? Even if it can, an
escort satellite is only good for protecting
one satellite in one orbit. Moreover, the
United States does not even have adequate
situational awareness to support passive
defenses, which is a less difficult challenge
than for active defenses.
Notwithstanding all these challenges,
in considering options for active defense,
there are missteps that can make the stability
problem even worse—such as putting
weapons in space. It is questionable how
weapons in space would support deterrence
given the limitations of what an armed
satellite could do to defend other satellites
or even itself. Space is an offense-dominant
environment—it is easier to attack assets
there than to defend them—so stability
would be a problem. If the United States
put weapons in space, China might do so
too; then, in a crisis, each side would have
an incentive to shoot first. Finally, violating
the taboo about space weaponization that
has emerged over the decades would be
30
politically costly for the United States in the
international community. Considering all
of these factors and the high cost of space
lift, the risks and costs of putting weapons in
space would almost certainly outweigh any
prospective benefits.
Norms. Norms have some effect at
low levels of confrontation and conflict,
but when push comes to shove and
the fighting begins, norms alone will
not suffice. However, they have a role
in strengthening first-strike stability in space
by both reducing benefits and raising costs.
What Does the United States Do?
In developing an answer to this question,
start with some basic principles. First, it
is important to recognize that it is usually
not productive to speak generally about
space deterrence. Rather, the United States
must identify which kinds of attacks on
which portions of which space systems
and under which circumstances it is trying
to prevent.
Second, the United States cannot deter all
kinds of attacks in space. While reversibleeffects attacks have so little cost associated
with them that they are likely not deterrable,
kinetic attacks are a reasonable threshold
for general deterrence, at least until the war
escalates.
Third, do not think in the stovepipes of
space deterrence, nuclear deterrence, cyber
deterrence, etc. The Chinese have a more
holistic view of deterrence than the United
States does. These facets of deterrence are
interdependent and should be considered
together.
Finally, the United States should stop
sending mixed messages to the rest of the
world. For example, it says that it does not
want to see attacks in space, but in unclassified
national space policy and military doctrine,
Cross-Domain Deterrence in U.S.–China Strategy
it talks about space control and dominating
space in a manner that implies that it thinks
fighting in space and warfare in space are
acceptable.
With these principles in mind, following
are recommendations on space policy,
strengthening the credibility of threats of
punishment and increasing U.S. ability to
deny benefits.
Space Policy. To strengthen deterrence
and first-strike stability, begin with a topdown coherent national space policy. Start
with explicitly condemning the use of force
in space. Declare that if others attack U.S.
space capabilities, the United States will
respond in ways, times, and manners of its
choosing. Avoid unclassified references to
space control and force application missions
in space. These policies would reinforce
norms against space attack by undermining
adversary efforts to depict the United States
as seeking to dominate space or develop
offensive space capabilities. They would also
provide a vague general threat in declaratory
policy while shaping the international
environment to make specific threats more
credible and build international support for
carrying them out, if necessary.
Make Threats of Punishment More
Credible. As the space warfare taboo
strengthens, threats to punish its violation
become more credible. Thus, U.S. policy
should focus on strengthening the taboo
and making punishment for its violation
more expected, more acceptable, and more
feasible.
Thus, U.S. policies and statements should
aim to shape international perceptions by
conditioning the international community
to accept the justice of the United States
punishing aggressors and conditioning
potential adversaries to take seriously
threats of terrestrial punishment in return for
attacks on space systems. In addition, the
United States should be willing to engage in
“brinkmanship” with adversaries in a crisis.
This involves fashioning threats of terrestrial
punishment (diplomatic, economic, and
military) in ways that put the onus of avoiding
catastrophic escalation on the adversary’s
shoulders. Finally, the United States should
develop means of exacting retributive costs
on the enemy’s space systems. Although titfor-tat is unlikely to work to U.S. advantage,
enemies must understand that they too will
suffer costs in space if they strike first.
Increase Ability to Deny Benefits.
This, too, involves managing perceptions.
The United States should emphasize the
resilience of its orbital infrastructure and
alternative capabilities while never divulging
specific vulnerabilities. It should make clear
to adversaries that, if worst comes to worst,
the United States will fight through regardless
of what happens to its space infrastructure.
Beyond that, the United States can make
its space capabilities more robust over
time by investing more in passive defenses;
considering subsidizing passive defenses on
commercial satellites; continuing research
on active defenses to identify technological
breakthroughs; dispersing space capabilities
across a larger number of platforms (as in the
GPS) and away from capabilities provided by
single, high-value satellites; making satellite
manufacture and space lift more responsive
to replenish losses more quickly; developing
terrestrial backups to space support wherever
possible; and exploring stealth, concealment,
and deception concepts to complicate the
adversary’s targeting problem.
B.Deterring Chinese Space
Attacks: Key Questions and
Participants’ Responses
This
section
presents
participants’
perspectives on the agenda questions
associated with the workshop session titled
31
Deterring Chinese Attacks on U.S. Space Capabilities
“Deterring Chinese Space Attacks.” The two
sources of information for this section are (1)
group discussions conducted immediately
after the presentation and (2) computer
inputs provided at the end of the session.
10.Can the United States effectively
dissuade space attacks without
demonstrating willingness to retal­
iate (i.e., are protection strategies
and establishing behavioral norms
sufficient for preventing space
attacks)?
There was a broad divergence of opinion
on the answer to this question. Respondents
were divided between no, yes, maybe,
and even retaliation might not be enough.
Arguments for no were:
• No, the United States needs to do both,
and, if necessary, it may need to actually
retaliate.
• No, because countries will place their
perceived interests in a crisis or war as
being more important than adhering to a
norm of this sort, especially considering
that an attack on a satellite would
not kill anyone. Moreover, in a highintensity conflict, it is very much in a
U.S. adversary’s interest to attack U.S.
satellites, and the United States will
probably be hitting whichever targets it
plans to hit anyway, so U.S. enemies will
attack U.S. satellites if they can.
Arguments for yes were:
• A modicum of deterrence (even as much
as a hint that U.S. behavior would become
erratic if it lost the information assurance
that space capabilities provide) coupled
with a set of alternatives (unmanned
aerial vehicles) that substitute for much of
the space capability might work.
• Yes, through the ability to impose political
32
costs on core values of the Chinese
Communist Party leadership.
• The United States can deter kinetic
attacks, and unlike other domains,
international norms may have some
impact—there is somewhat of a taboo
about space warfare. Moral superiority
in any conflict with the United States
will be important, and so if countries
will really unite and react strongly, it
is possible that this will impact the
Chinese calculus.
An argument for maybe was:
• Possibly, but it would not only require
reducing the vulnerability of U.S. space
systems and establishing norms but also
reducing U.S. reliance on space systems
to make such attacks less attractive in the
first place.
Finally, several respondents raised the
point that even retaliation might not suffice:
• Demonstrating willingness to retaliate
might not be enough either, unless
the other side is at least somewhere
close to as reliant on space systems
as the United States. If they are much
less reliant on space than the United
States (even if they are roughly equally
reliant overall but less so in a particular
scenario), they might conclude that the
benefits of attacking U.S. space systems
would outweigh the costs of whatever
damage the United States might inflict on
theirs.
• Unless U.S. systems are invulnerable,
or unless the United States does not
depend on space for military, economic,
and other purposes (i.e., space systems
are irrelevant), it cannot dissuade space
attacks. The ability to retaliate is less
relevant. Also, it is not clear that China
is becoming more dependent on space,
or more precisely, that it will become as
Cross-Domain Deterrence in U.S.–China Strategy
dependent or comparably dependent on
space as the United States.
A number of respondents emphasized
the importance of context for any attack on
U.S. space assets, which is likely to involve a
broader U.S.–China conflict:
• It is hard to conceive of attacks on space
assets occurring outside the context of a
broader conflict.
• Behavioral norms will have almost no
deterrent effect on China’s willingness to
launch space attacks. Protection strategies,
on the other hand, seem more promising,
although attacking space assets will
probably always be cheaper and easier
than defending them. In other words,
deterrence by denial would be great,
but it is hard to be optimistic. That leaves
deterrence by punishment, which faces its
own problems, ranging from a lack of key
Chinese space capabilities to difficulty of
identifying and hitting comparable targets.
On the other hand, if China is attacking
U.S. satellites, then the United States is
almost certainly already engaged in a full
conventional conflict, which means that it
is already hitting most strategic targets. So,
in the end, it is quite pessimistic to think
of the U.S. ability to deter space attacks
in scenarios in which the Chinese would
actually contemplate doing so.
11.Are there effective retaliatory
threats that stay within the space
domain, or must retaliatory threats
cross domains to be effective?
Asymmetries between the United States and
China regarding dependency on space were
noted by many respondents, leading to the
judgment that effective retaliatory threats
must cross domains:
• Asymmetries in how the United States
and China define and perceive the
space domain are important here. It
looks as though the Chinese take a more
expansive view of the domain than the
United States.
• Given the asymmetries in dependence
on space support between the United
States and China, it is hard to think of
any retaliatory threats that stay within
the space domain that would be
effective in deterring attacks on U.S.
space systems. China’s potential future
reliance on satellites to locate and target
U.S. carrier task forces in the DF-21
era might hold a possibility for holding
Chinese reconnaissance and ocean
surveillance satellites at risk to deter
them from attacking U.S. corresponding
assets; however, it is likely that U.S.
Navy and Joint commanders will argue
that the United States will need to
target China’s satellites to interdict that
kill chain regardless of whether the
Chinese exercise initial restraint. With
this in mind, it will be difficult, perhaps
impossible, for U.S. leaders to make
credible assurances of restraint.
•Given asymmetries of dependence,
capability, and roles, it is unclear what
space retaliatory measures would be.
• Focused on China and on conflicts in
the western Pacific, the asymmetries in
dependencies in space are so profound
that any credible threat the United
States could pose in the space domain
would be insignificant compared to the
cost of the loss of U.S. space assets. The
only effective threats would be across
domains to find targets of comparable
value to hold at risk. The challenge
there, of course, is establishing the
necessary precondition to an effective
deterrent: a clear path of escalation
unconditionally set in the mind of
the adversary.
33
Deterring Chinese Attacks on U.S. Space Capabilities
• Must be cross-domain. The United States
depends more on space than anyone else
does, so space versus space retaliation
and counter-retaliation hurts it more
than them.
• Because of asymmetries, a retaliation
strategy (at this point) that does not cross
domains is unlikely to be very dissuasive.
• One of the key challenges in crossdomain discussions is the dialogue
about the relationships between one
domain and all the other domains. In
the strategic view, a space attack might
require the United States to retaliate
in the most appropriate domain. True
cross-domain
strategic
approaches
should require the National Command
Authority to select the most appropriate
domain or combination of domains—
with the appropriate proportionate or
disproportionate response—to include
the timing of the response.
• It likely depends on how dependent the
adversary country is on space systems.
If it is highly dependent, effective
retaliatory threats could be in-kind. If
it is much less dependent on space,
the retaliatory threats would probably
have to cross domains so they could
threaten something of greater value to
the adversary.
• The United States needs to think harder
about the idea of proportionality vis-àvis Chinese attacks against U.S. space
assets. Punishment probably needs to be
inflicted on non-space assets.
• I think the answer turns on one’s
assessment of China’s reliance on space
assets, in particular, the degree to which it
will or will not become more dependent
on space systems. The more that China
relies on space capabilities of its own,
the more effective U.S. retaliatory threats
that stay within the space domain will
34
be. Otherwise, cross-domain retaliatory
threats seem the only option, even if that
remains a poor one.
12.What principles should guide the
development of U.S. cross-domain
deterrence policies for space
(e.g., are red lines useful, and, if
so, what should they be and how
should they be communicated)?
Several respondents focused on the potential
for, and problems with, red lines:
• The United States probably cannot deter
things like dazzling and jamming, but
maybe it could establish a red line that
prohibits destructive attacks (those that
generate debris in space) or a red line
that prohibits attacks against certain
types of satellites (such as missile early
warning/launch detection satellites).
• Red lines might make more sense here
than in the cyber domain because of
the seemingly greater clarity in type
and nature of attacks. However, any
circumstance where China is conducting
anti-satellite attacks will already (or
immediately) involve the United States
conducting extensive conventional and
cyber operations—so the United States
will be well beyond cross-domain
deterrence and into cross-domain
warfighting.
• The United States may be able to deter
attacks on its satellites during peacetime/
crisis/limited skirmish by stating that
it will treat an attack on U.S. satellites
as an attack on the U.S. homeland. The
United States cannot deter attacks on
its satellites during high-intensity war
unless it is willing to go nuclear.
Other respondents focused on damagelimitation capabilities such as protection,
redundancy, and mitigation:
Cross-Domain Deterrence in U.S.–China Strategy
• The United States needs to improve
its non-space capabilities for various
functions provided by satellites so as
to reduce the payoff to an adversary
from attacking U.S. space assets. Better
non-space capabilities may be the most
important single step that the United
States could take.
• Space systems are inherently vulnerable
in wartime, and the United States does
not want to be in a position where it has to
escalate beyond standard conventional
war to keep enemies from attacking
U.S. satellites. A more robust strategy
is to replace satellite capabilities with
unmanned aerial vehicle and related
radio frequency capabilities.
One respondent focused on the problems
of attribution and shared space assets:
• Assigning attribution to acts in space is
more difficult than people think. If an
attack occurs without the generation of
debris (or even if debris is generated), it
is often difficult to determine attribution
with certainty. The element of deniability
makes retaliation more costly in the
international community. The best option
for the United States is to conduct its
own retaliatory attacks quietly, preferably
disabling Chinese space (and possibly
other) assets without obvious kinetic
strikes.
The problem becomes more complicated
because many space assets are
shared: what should the reaction be
if a commercial satellite is attacked,
especially if the attack is reversible and
attribution is not certain? The United
States should explore policy options
that include improving its capabilities to
attribute sources of attacks (of all natures:
kinetic and non-kinetic) and identify
what nations share use of commercial
satellites with it (e.g., European Union,
Canada, India, etc.).
Finally, the impact on norms of behavior
in conflict that extends to space should
be considered beyond simply United
States versus China. Russia and emerging
space powers such as India (both of
which are neighbors to China and are
potential adversaries) should be brought
into the discussion early. Chinese losses
in space would put them behind their
neighbors, and this prospect should deter
their willingness to risk losses to space
assets at the hands of the United States,
especially because their problems with
attribution are even more severe.
• The United States should emphasize the
ability to fight through a counter-space
attack, but if the attack in question truly
undermines U.S. conventional power,
U.S. stakes in the conflict will have
dramatically increased.
• There is definitely a need for enhanced
resilience and bolstering norms against
space attacks.
One respondent emphasized that
deterrence relies on identifying and holding
at risk assets in other domains that the
Chinese value:
• The United States needs to identify
Chinese assets that are as important to
China as space assets are to the United
States (surveillance systems to locate
and track U.S. aircraft carriers are one
possible example). Holding these assets at
risk could form the basis for cross-domain
deterrence. However, for such a strategy
to be effective, the United States should
(1) be able to signal, in a crisis, that it is
capable of holding such assets at risk and
(2) be willing not to strike these assets if
China does not attack space assets.
35
5
DETERRING CHINESE
NUCLEAR USE
While nuclear use by China or the United
States against the other may seem far less
likely than other threats discussed during
this workshop, the physical, social, and
psychological consequences of any such
use would, in many scenarios, dwarf those
due to warfare in other domains. However
remote, the possibility of escalation to
nuclear war provides a backdrop to even
minor confrontations. Also, should a severe
crisis emerge, both the United States and
China recognize the need to deter nuclear
use and counter nuclear threats by the other.
At the same time, the United States and
China have significantly different policies,
strategies, and forces. Lack of appreciation
of these differences and the reasons for them
can undermine stability by encouraging
mirror imaging and other forms of erroneous
thinking.
Section A summarizes the presentation on
deterring Chinese nuclear use provided to
the workshop, followed by Section B on key
questions and participant responses.
A.Presentation Summary:
Deterring Nuclear Threats
in a U.S.–China Crisis or
Conflict
by
Michael Chase
Sources for this presentation include
the following: (1) for force modernization
information,
unclassified
authoritative
U.S. intelligence community assessments
such as the Department of Defense China
military power reports,15 National Air
and Space Intelligence Center ballistic
and cruise missile reports,16 and Office of
Naval Intelligence People’s Liberation Army
Navy reports;17 and (2) for policy, strategy,
and doctrine, Chinese language books
such as the The Science of Strategy, SSAC,
and Intimidation Warfare; articles from
journals such as China Military Science and
Military Art; and selections from Second
Artillery newspaper and other Chinese
military newspapers.
U.S. Department of Defense, Annual Report to Congress: Military and Security Developments
Involving the People’s Republic of China, 2013; available at www.defense.gov/pubs/2013_china_
report_final.pdf.
15
16
National Air and Space Intelligence Center, Ballistic & Cruise Missile Threat, NASIC-1031-0985-13,
2013; available at http://www.afisr.af.mil/shared/media/document/AFD-130710-054.pdf.
Office of Naval Intelligence, The People’s Liberation Army Navy: A Modern Navy with Chinese
Characteristics, 2009.
17
37
Deterring Chinese Nuclear Use
It is important to recognize the limitations
of Chinese doctrinal publications. First,
although they reveal how China’s military
plans to fight and what the United States
should look for in training and exercises, they
do not necessarily dictate how it will fight. In
any severe crisis, nuclear decision making
would ultimately fall to a very small number
of senior Chinese Communist Party leaders,
many of whom probably never thought much
about the details of nuclear warfighting. In
addition, changes in China’s external security
environment could erode its traditional logic of
minimum deterrence, reducing the relevance
of some current writings. For example,
while the United States is currently the most
important component in Chinese nuclear
calculus, certain events could make regional
nuclear deterrence rise on China’s list of
priorities. Significant problems could develop
in China’s relations with Russia or India, or
Pakistani actions might precipitate an Indian
response that China views as threatening to
its own interests.
Policy
Chinese strategists see nuclear weapons
as useful primarily for deterring nuclear
attacks or countering nuclear coercion.
Thus, China has maintained a No First Use
policy since its first nuclear test in October
1964; contemporary Chinese doctrinal
publications continue to adhere to No First
Use. The People’s Liberation Army Second
Artillery Corps emphasizes survivability of
nuclear forces because it is presumed to
be extremely likely these forces will be
ordered to launch unless they have been
already attacked.
The SSAC suggests that nuclear deterrence
could also counter conventional strikes against
strategic targets. Higher-intensity deterrence
measures that lower the nuclear threshold
could invoke more explicit nuclear threats
to ensure that enemies clearly understand
the risks they have exposed themselves to
or could adjust the No First Use policy by
conducting a nuclear demonstration or a
smaller-scale nuclear strike. Although it
is unclear whether China would launch a
nuclear strike in response to a conventional
attack, SSAC mentions four conditions under
which this might happen:
1.When the enemy threatens to launch
conventional strikes against China’s nuclear
facilities or nuclear power stations;
2. When the enemy threatens to attack major
strategic targets like hydroelectric power
stations (such as Three Gorges);
3. When the enemy threatens to carry out
attacks against the capital, major cities, or
other political or economic centers; or
4. When China faces impending disaster
because it is losing a high-stakes
conventional conflict.
Whichever higher intensity deterrence
measures are taken, their purpose would
be to frighten the enemy into stopping its
conventional attacks or, at least, to reduce the
effectiveness of their attacks. Such measures
could only be undertaken under the authority
of the top-level political leadership of China.
There is some discussion in the literature that
higher-intensity deterrence actions could
trigger escalation, rather than deter it, but not
a lot of in-depth discussion of how to avoid
this outcome or which factors would hinder
or contribute to such an occurrence.
Information Office of the State Council, People’s Republic of China, China’s National Defense in 2006;
available at http://www.fas.org/nuke/guide/china/doctrine/wp2006.html#0.
18
38
Cross-Domain Deterrence in U.S.–China Strategy
Strategy and Posture
Chinese analysts do not describe their
posture in official documents as “minimum
deterrence,” but rather as “a lean and
effective” nuclear force that meets China’s
national security needs.18 Contemporary
Chinese authors also refer to “minimum,”
“maximum,” and “medium” strength nuclear
deterrence. A minimum strength nuclear
deterrent would threaten a handful of secure
nuclear retaliatory weapons; a maximum
strength nuclear deterrent characterizes
the U.S.–Soviet relationship during the
Cold War, involving thousands of nuclear
retaliatory weapons. China is transitioning
from a minimum strength toward a medium
strength nuclear deterrent, a much more
substantial force than it had in the past, but
still much less than that of Russia or the
United States.
Similar to U.S. nuclear strategy, the
primary goal of China’s nuclear strategy
is to “deter other countries from using or
threatening to use nuclear weapons against
China.”19 Chinese nuclear policy emphasizes
“no first use of nuclear weapons at any time
and under any circumstances,” although
exceptions for certain conventional attacks
are possible.
China’s nuclear force modernization
is aimed at creating a “lean and effective”
nuclear deterrent that meets China’s national
security needs. The number of weapons
China views as appropriate is driven by
evolving Chinese perceptions of its threat
environment, including threats to its nuclear
deterrent from both conventional and
nuclear attacks as well as missile defenses.
Also, China does not want an excessive
number of nuclear weapons. At best, an
19
excess is a waste of resources; at worst,
it could be destabilizing and undermine
China’s security.
Beijing’s current emphasis is on
enhancing the survivability and firepower of
its nuclear force by deploying road-mobile
intercontinental ballistic missiles (ICBMs)
and nuclear-powered ballistic missile
submarines (SSBNs). To develop these
systems, China is pursuing a number of
enabling technologies, including warhead
miniaturization (for smaller, mobile missiles
and MIRVed systems); improved ability
to penetrate missile defenses; improved
operational command, control, and support
systems; and improved accuracy, as well
as increasing the number of missiles and
warheads. With these new systems, China’s
nuclear forces will eventually be composed
of a road-mobile ICBM (some MIRVed),
a silo-based ICBM, and a sea-based strategic
deterrent.
China has traditionally maintained its
nuclear forces at an extremely low level
of readiness. This is likely changing with
China’s emphasis on survivability and the
transition to road-mobile ICBMs and SSBNs.
Some units will regularly be at high levels of
readiness, while others will not. The Chinese
apparently practice moving from one level
of readiness to another.
China’s emphasis on a secure secondstrike capability could contribute to
greater stability in the U.S.–China strategic
relationship because there would not be
an incentive for either side to launch a
large-scale strike. As China increases the
survivability of its nuclear force, the United
States will come to realize that any attempt
at a damage limitation strike would not be
Ibid.
39
Deterring Chinese Nuclear Use
in its interest because China will still be
able to retaliate with a significant number of
nuclear weapons. China will become more
of a medium nuclear power, smaller than
Russia and the United States, but larger than
it is today. As such, China could complicate
future arms control negotiations, and aspects
of Chinese doctrine could heighten escalation
risks in a crisis or conflict.
Nuclear Counter-Strike Campaigns
Writings emphasize that any nuclear counterstrike campaign would be strictly controlled
by the Supreme Command. It could be
conducted solely by the People’s Liberation
Army Second Artillery Corps or jointly with
the People’s Liberation Army Navy. The
People’s Liberation Army Air Force does not
have or want a nuclear mission at this time.
The objectives of a nuclear counter-strike
campaign would be principally aimed at deescalation or termination of a conflict.
What would China strike? Chinese writings
do not discuss striking what the United States
would call counter-force targets such as ICBM
fields. However, they also do not emphasize
pure counter-value (i.e., cities) targeting.
They consider a wide range of both military
and civilian targets, including military bases
and command centers, communications and
transportation hubs, political and economic
centers, important industrial facilities, and
other strategic and campaign-level targets.
One change from the past is that there
seems to be less discussion about striking
concentrations of forces. Older statements
were probably based on concerns about
massing Soviet troops, but China is no longer
as worried about ground invasions.
Chinese writings also do not emphasize
electromagnetic pulse (EMP) strikes, but
such strikes would be consistent with
China’s strategic-level political objectives
of changing the enemy’s decision calculus,
40
in addition to the physical effects of
such attacks.
B.Deterring Chinese Nuclear
Threats: Key Questions and
Participants’ Responses
This
section
presents
participants’
perspectives on the agenda questions
associated with the workshop session titled
“Deterring Chinese Nuclear Use.” The two
sources of information for this section are
(1) group discussions conducted immediately
after the briefing and (2) computer inputs
provided at the end of the session.
13. Does China share the U.S. goal of
isolating the nuclear domain?
Some respondents questioned the premise of
this question—that the United States truly has
the goal, or takes it seriously in practice—of
isolating the nuclear domain:
• Rhetorically, both the United States and
China seek to isolate the nuclear domain
(China declares no first use; the United
States seeks to create the conditions to
allow sole purpose). Neither commitment
is entirely credible over the long term. At
the operational level, neither side isolates
the nuclear domain. The United States
has dual-use B-2s, B-52s, early warning,
etc. China has shared C2, missiles with
conventional and nuclear variants, etc.
• The United States does not explicitly seek
to isolate the nuclear domain as a matter
of policy, but it has stated that it wants to
move toward a policy that de-emphasizes
nuclear weapons and relegates them to
the sole purpose of deterring nuclear
threats; the United States clearly states
that it is not there yet. It still considers
nuclear weapons a viable deterrent to
some types of conventional or other
non-nuclear attacks. The United States
Cross-Domain Deterrence in U.S.–China Strategy
blurs the line with missile early warning
satellites, B-2s, etc.; China also blurs
the line, at least to some extent, with a
strategic missile force that fields nuclear
and conventional systems. China seeks to
mostly isolate the nuclear domain with
its No First Use policy but also to deter
at least some conventional attacks with
strategic effects.
Other respondents were fairly evenly
divided between answering in the negative
and expressing uncertainty. Those who
answered no pointed to the role of nuclear
weapons in countering U.S. conventional
superiority:
• No, it is not in China’s interest to share
this goal. Its nuclear force represents a
strong counter not only to U.S. nuclear
intimidation/coercion but also to U.S.
advanced conventional weapons.
• No, China does not share the goal of
isolating the nuclear domain today, at
least not in a high-stakes conventional
conflict where China is losing. In such
a circumstance, China will see huge
incentives to escalate to the use of nuclear
weapons to coerce the United States to
halt combat operations before China
suffers the consequences of conventional
defeat. In short, whereas the United
States would like to fight and win its
wars at the conventional level (because
it possesses conventional superiority),
China will see nuclear escalation as a
tempting strategy for compensating for its
conventional weakness in a high-stakes
war. The United States may want a clear
firebreak, but it faces different strategic
circumstances. Rhetorically, China may
claim to want to isolate (and eventually
eliminate) nuclear weapons, but unless
China can reverse its conventional
inferiority, there is not much stock in
that rhetoric.
The respondents who emphasized
uncertainty offered the following points:
• I don’t know. Chinese goals can change
without notice, depending on the views
of the top leadership at any given time.
• I am not at all sure that this is true for
China; maybe rhetorically, but not
operationally or in practice.
14. What U.S. actions in other domains
could provoke Chinese nuclear
threats or use?
All respondents mentioned large-scale
conventional attacks, especially those
threatening Chinese Communist Party
legitimacy or China’s nuclear deterrent
capability:
• Attacks on C4ISR, critical infrastructure,
large-scale counter-value attacks, e.g.,
Tokyo 1945, Taiwanese secession or
Tibet, and insurgency support in Xinjiang
and Tibet.
• U.S. actions include conventional or
possibly cyber attacks against nuclear
forces or associated command, control,
and communications systems; those that
convey the impression the United States
seeks to overthrow the Communist Party,
kill the party leadership, or otherwise
pursue regime change; those leading to
a catastrophic military defeat that could
gravely imperil domestic stability or could
potentially lead to the collapse of the
regime; or perhaps more likely, actions
that are misinterpreted or misunderstood
by the Chinese as being aimed at
these objectives or lead to inadvertent
escalation through some other type of
miscalculation.
• Jeopardizing Chinese Communist Party
authority/survival would probably raise
questions about nuclear use. This may
41
Deterring Chinese Nuclear Use
not be limited to U.S. nuclear use but
also U.S. support for other states/entities
that jeopardizes Chinese Communist
Party authority/survival.
• Large-scale conventional attacks against
targets in China, especially targets
deep inland or associated with national
leadership, could easily cause escalation
to the nuclear level.
• U.S. actions could include kinetic attacks
against Chinese strategic or leadership
targets or possible non-kinetic attacks
against nuclear C2.
• Looming Chinese defeat in a high-stakes
conventional war could trigger nuclear
escalation as a means of coercing a
military stalemate that would forestall
that outcome.
•Actions include U.S. conventional
attacks against Beijing, Chinese nuclear
forces, or Chinese nuclear power.
Another alternative is a large-scale U.S.
conventional defeat of Chinese forces
that risks the legitimacy of the Chinese
Communist Party.
15.What Chinese actions in other
domains could provoke U.S.
nuclear threats or use?
Analogous to some answers to the previous
question, respondents identified scenarios
involving Chinese first nuclear use and
impending catastrophic U.S. conventional
defeat. Some responses suggest that
U.S. concern with the survivability and/
or legitimacy of its own leadership is a
concern that parallels that of the Chinese
Communist Party:
• Same as Question No. 14, and also
consider similar attacks on these targets of U.S. allies. For all these questions, there are the risks of unintended or
42
unforeseen
consequences,
miscal­
culation, inadvertent escalation, and the
fog of war that must be reckoned with.
• Potential Chinese actions include those
that lead to a catastrophic or humiliating
defeat that could shatter U.S. credibility,
gravely undermine U.S. alliances,
and severely diminish U.S. regional
influence and interests—or perhaps more
likely, actions that are misinterpreted or
misunderstood as being aimed at these
objectives or that lead to inadvertent
escalation through some other type
of miscalculation.
• Chinese actions that precipitated massive
U.S. casualties or existential survival
questions would provoke U.S. nuclear
threats. Chinese use of nuclear weapons
would precipitate U.S. nuclear threats/use.
• What impact would sinking a U.S. carrier
battle group have?
• The Defense Science Board suggests
a sufficiently severe, high-casualty
event might instigate the United
States. Overrunning a U.S. ally (e.g.,
Republic of Korea) might also have the
same effect.
• Other actions include large-scale Chinese
conventional defeat of U.S. forces (e.g.,
multiple carriers sunk).
• Any action that would result in damage
and casualties on par with a nuclear
attack could provoke a U.S. nuclear
threat or use.
Some respondents did not believe there
were plausible scenarios in which the
Chinese could provoke U.S. nuclear use:
• There are not any foreseeable, plausible
Chinese actions in the non-nuclear
domain that would provoke U.S. nuclear
threats or use.
Cross-Domain Deterrence in U.S.–China Strategy
• It is hard to see how any credible Chinese
conventional/cyber attacks would be an
existential threat to the United States, so
it is unlikely that the United States would
go nuclear. Moreover, China is unlikely
to pursue any strategy aimed at inflicting
huge civilian casualties or massive
damage to U.S. infrastructure through the
use of conventional weapons. The United
States will not use nuclear weapons to
keep China from conquering Taiwan, and
it may not even be willing to conduct
large-scale conventional strikes against
targets in China.
• This is not likely. The United States has
a strong sense of the nuclear taboo, and
since the end of the Cold War, it is not
apparent that use of nuclear weapons is
part of national strategy.
16.
Do Chinese authors predict
that their own or U.S. precision
conventional, cyber, or space
attacks will elicit a nuclear
response?
This question differs from the previous two
questions in that it focuses on Chinese
writings. Respondents indicated significant
uncertainty in their answers:
• In some cases, yes, depending on the
author involved and the targets attacked
by the United States. It is unseen whether
they predict U.S. nuclear response in
public, though one obviously cannot be
sure in private.
•People’s Liberation Army writings
indicate that combined deterrence
is likely to lead to deterrence of U.S.
nuclear use. It is unclear what would
elicit a nuclear response.
• It is not likely. China very rarely
considers and accurately determines
how its actions are interpreted by others.
It views its actions as defensive and
therefore not provocative.
17. Should the United States explicitly
retain or relinquish the option
to respond to cyber and space
attacks with nuclear weapons,
cultivate ambiguity, or be silent on
the matter?
Almost all respondents opted for ambiguity
or silence, some noting that silence will
foster ambiguity:
•The United States should cultivate
ambiguity to make China see that it is
potentially playing with real fire.
• Silence is golden. It breeds ambiguity.
• The United States should be silent about
possibilities of responding to cyber or
space attacks with nuclear weapons. It
should not make any explicit threats that
lack credibility, and it is inconceivable
that the United States could convince
anyone that it would resort to nuclear
weapons against any state, much less
a nuclear-armed adversary, in reprisal
for attacks on uninhabited machines in
space or U.S. computer networks. The
only way such attacks could lead to a
U.S. nuclear response would be if they
degraded U.S. warfighting capability
so severely that the United States
was on the verge of an unacceptably
catastrophic military defeat (whatever
that might be). However, even in those
circumstances, it would be events
in the terrestrial domain that would
lead to the risk of nuclear escalation,
not the space or cyber attacks in
isolation. Therefore, threatening a
nuclear response in efforts to deter
space and cyber attacks would not
be credible.
43
Deterring Chinese Nuclear Use
• Be silent on the matter. Making threats
that the United States cannot make
credible only propagates the impression
of U.S. fear of such attacks, potentially
increasing the probability an opponent
would carry them out. Silence will result
in ambiguity.
•The United States should cultivate
ambiguity or be silent on the matter, but
if this is about rhetorical policy, then this
issue does not matter much.
44
One respondent suggested relinquishing
the option to respond to space and cyber
attacks with nuclear weapons:
• Everything depends on whether it is
U.S. policy to reduce the number of
situations/contexts in which nuclear use
is contemplated. If so, the answer is to
relinquish the option because any cyber
attack implicates the systems’ owners
as well as the attacker, and there are no
space assets that could not be replaced
within a few years.
6
FUTURE RESEARCH
In the final workshop session, three questions
were posed with the intention of identifying
future research needs and productive research
approaches. Each participant was asked
to verbally address one (or more) of these
questions and to provide any additional input
using the computer.
18.What else does the United States
need to understand about China?
Several respondents focused on understanding
central aspects of China’s strategic culture—
its role in the international system, its values,
and its fears:
• The United States needs to understand
that China sees itself as a status quo power
and how it defines and defends the status
quo as it sees it. Both the United States
and China may see themselves protecting
the status quo but find the other’s actions
as offensive. History shows that wars are
not caused by rising powers threatening
international systems but by declining
powers unwilling to relinquish their
positions gracefully.
• The United States needs a better sense
of what the Chinese Communist Party
values because deterrence can succeed
only when it can hold at risk what
China values. The United States assumes
it is China’s nuclear capabilities and
thinks it might be China’s economic
advances. The United States canonically
assumes population is most important,
but maybe it is not. It emphasizes the
Chinese Communist Party survival/hold
on power. However, how does one go
about threatening that? Defeating them
in war? Eliminating the People’s Armed
Police (PAP) database? Destroying cities?
Propaganda?
• At levels of violence lower than holding
at risk the Chinese Communist Party’s
grip on power, the United States needs
to identify China’s fears. What assets
does China most fear the United States
will hold at risk in a crisis (equivalent to
U.S. concerns about its space assets). If
those fears are known, then the United
States can take advantage of them and
enhance deterrence.
Other respondents focused more directly
on deterrence and compellence:
• The United States needs to understand
everything about China’s approach to
nuclear warfare and deterrence from its
reference point, rather than in the context
of U.S. deterrence theory. Mirror imaging
must be eliminated. The United States
and China do not think of deterrence in
the same way. The United States must not
presume that China thinks about strategy,
politics, escalation, and information
warfare and operation the way the United
States does.
• The United States needs to better
understand deterrence of the strong by
the weak. This requires a much more
rigorous examination of the use of
nuclear weapons to deter conventionally
45
Future Research
superior adversaries, whether locally
or globally.
• The United States needs to separate
peacetime deterrence of the use of
force and deterring further escalation
during wartime or limited conflict.
How does China see these differences?
What happens when you move beyond
Phase 0 or Phase 1 and need, in a limited
war, to deter hostilities from moving into
other domains? Under what conditions
would China capitulate rather than
escalate? There is a huge divergence in
opinions among people with different
backgrounds on the question of how you
would persuade the Chinese to decide to
discuss, rather than escalate, the conflict.
• Deterrence involves taking a course
of action to dissuade China from
action. The flip side of this is coercive
persuasion: how to persuade China to
act as you want it to act. Objectives
of coercive persuasion might include
stopping intrusive cyber surveillance and
stealing intellectual property; adopting
a much more conciliatory policy about
the territorial claims in the South China
Sea or East China Sea; and accepting
the reality of Taiwan’s existence as an
independent sovereign state within a
One-China framework. How does China
conduct coercive persuasion?
19. Are there specific scenarios that
should be examined?
Table-top exercises provide a primary
analysis tool to explore strategy, options,
crisis evolution, and decision making.
However, such exercises are only as good
as the scenarios on which they are based.
We have therefore posed this question to the
workshop to help shape an effective agenda
for research.
46
There was no shortage of participant
suggestions regarding scenarios. Many
suggestions involve combinations of actors
other than a direct confrontation between
the United States and China. Such scenarios
included those in which the United States is
not a principal actor; others involved China
and a U.S. ally or China and another regional
actor; and still others involved neither the
United States nor China as the principal
actors:
• With respect to scenarios, discussions so
far have pitched the United States against
China building to, or in, some type of
conflict. A more plausible scenario may
be one in which China has an altercation
with an ally of the United States, with the
United States acting as a third party. As the
altercation escalates, what role should the
United States take in deterring both parties
from further escalation?
and
China–Vietnam,
• China–India,
China–Russia are scenarios that must be
examined, not just Taiwan or Japan, and
we have not considered them.
• The United States needs to address
scenarios in which China is not
one of the main antagonists but
could be brought in anyway due to
Chinese equities; India–Pakistan or
U.S.–Pakistan would be the more
interesting of these.
• The United States should consider
scenarios in which U.S. allies use
or threaten to use their own military
capabilities.
• Consider scenarios in which there are
many players. Consider scenarios in
which other actors take the lead (e.g.,
India, Pakistan, North Korea, or Taiwan)
with a more nationalistic and/or more
provocative government.
Cross-Domain Deterrence in U.S.–China Strategy
• The United States needs to play multiactor scenarios.
• Mostly we address a direct Chinese
attack on a U.S. ally. Consider scenarios
such as India–Pakistan, in which China is
dragged into a crisis with someone else
in the region (not the usual direct attack
by China on an ally).
• Consider scenarios in which players other
than the United States and China, even
impoverished states, have significant
cyber capabilities. Such a cyber-capable
third party (e.g., North Korea) may have
caused an attack or at least the evidence
suggests so. How would that play out in
U.S.–China relations?
• What would happen in a Korean collapse
or Korean war?
Three respondents suggested scenarios in
which the internal situation in China is an
important driver:
• The United States should think about
scenarios in which China is not a
rising power but is either stagnant
or declining. This could be due to a
confluence of a number of its various
problems (economic, environmental,
demographic, etc.). China could see
its window of opportunity closing to
resolve in its favor some of its outstanding
territorial disputes, such as Taiwan. This
situation could motivate greater Chinese
reliance on nuclear weapons.
• The United States needs to play China
with internal dissent and instability and
examine how that affects the deterrence
equation and escalation dynamics.
• What if the Chinese decision-making
structure breaks down, where the Central
Committee is no longer in complete
control over the forces that are acting?
Two respondents made suggestions
on more realistically representing Chinese
doctrine and decision making:
• In most war games, Blue (the United
States) has all the bureaucratic problems
of reality. By contrast, although Red
(China) is at least as bureaucratized as
Blue, Red is usually played as without
any bureaucratic, internal political
divides. Thus, the United States needs to
play China with a bifurcated command
structure, a bureaucratized government,
provincial authorities’ inputs, and
concern for domestic popular opinion.
• Play a series of interactive scenarios,
doctrinally informed on both sides. This
could help us answer the question about
whether doctrine matters at all.
Two respondents suggested scenarios
involving a cyber attack that cause unintended
yet significant damage:
• Consider a significant cyber attack on
civil infrastructures in the United States
that leads to widespread civil damage
but limited direct human casualties. The
damages caused by such an attack could
have been inadvertent. What kind of
response would such an event provoke?
How would the United States react?
• What happens when there is a cyber attack
that inadvertently causes catastrophic
damage (e.g., one that blacks out half of
China or causes a catastrophic reaction at
a nuclear power plant)?
Two respondents focused on developing
and assessing options for responding in
various circumstances:
• Consider a crisis with China in which
some non-kinetic satellite attacks occur.
How does the United States respond
to best achieve its objectives? This is an
important part of deterrence that the
47
Future Research
United States is not exploring as much as
it should be. Deterrence is too divorced
and too separated from crisis response
and warfighting.
It might also want to engage Chinese
analysts in a discussion to examine
scenarios in which, for example, a natural
disaster in space occurs or a cyber attack
on critical infrastructure in the United
States or Europe, focusing on how the
United States and China communicate
to ensure stability during the period
of uncertainty as to what happened
and avoid inadvertent escalation. This
approach would allow discussions
with China, which is unwilling to talk
about these issues in an international
U.S.–China context.
• The United States needs to examine a
range of difficult scenarios involving
cross-domain attacks, in part to develop
credible military and diplomatic options
to draw from in a real crisis. As an
example, consider a crisis that has not
escalated, but China starts blinding and
jamming U.S. satellites. What do we do?
Credible deterrence must be anchored in
credible plans and capabilities.
Two respondents suggested scenarios
that explore the reaction to nuclear use:
• It seems that thinking stops as soon as
there is a nuclear detonation. What
happens after a nuclear detonation
occurs? The United States needs to
discuss ways to de-escalate the situation
after the detonation of a nuclear weapon,
for example, near a carrier battle group
or at a military facility.
• What if North Korea used nuclear
weapons? If North Korea killed a million
people, the United States would be
unlikely to settle for anything less than a
regime change. How would China react?
48
China has been generally cautious,
probably believing time is on its side if
nothing throws a monkey wrench into
the works. China’s GDP is growing fast,
as is its standard of living. The economic
situation in China is improving compared
to other countries, even if it is not yet
where it wants to be. Thus, China does
not want such a crisis but could be
unwillingly involved in a conflict started
by North Korea or involving India–
Pakistan.
The remaining suggestions do not lend
themselves to grouping, other than to note
that many are among the least well-examined
possibilities:
• Consider lower-end scenarios in which
there is coercive persuasion but not
enough use of force to ensure a military
response with the potential of the situation
escalating to military conflict. It is easy
enough to envision this for a China–Japan
crisis. Also South China Sea scenarios
in earlier phases could escalate from
Chinese use of economic pressure and
maritime law enforcement capabilities
(e.g., against the Philippines) to a level
at which China might contemplate using
military force. Nuclear deterrence would
not be relevant at that point; what kinds
of conventional, cyber, and space actions
or activities could result?
• Consider situations that occur in locations
more geographically distant from China
that would force the Chinese into greater
(than the East China Seas or a Taiwan
scenario) reliance on space systems,
which could at least partially change
some of the dynamics. The stakes would
not be as high, nor would conflict be at
the same level.
• The United States could also explore
scenarios involving China that do not
Cross-Domain Deterrence in U.S.–China Strategy
arise out of a conflict. For example,
how do the United States and China
maintain stability after a natural disaster
in space degrades some satellites?
The United States has not considered
how civilian leaders will think about
these catastrophes in a crisis. Do they
understand cross-domain relationships,
and how might their perceptions differ
from the military? Civilian Chinese
leaders may be more concerned about
economic interests in space than
military leaders, who may be more
concerned with tactical advantages that
stem from less military dependence on
space assets.
• The conventional wisdom that disparity
in nuclear capability does not matter
much in a crisis has not been adequately
challenged. Thus, the United States
needs to analyze scenarios that explore
the importance (or not) of parity in
nuclear capabilities. Unfortunately,
because it is mainly those who subscribe
to the conventional wisdom that would
play out the scenarios, it is likely to
reflect this conventional wisdom in
scenario-based exercises. We have to
think of ways to reduce the likelihood of
this happening.
20. What has not been considered?
A number of responses to this question
echoed scenario suggestions for the previous
question.
• The United States needs to consider
options to become less dependent on
space. In particular, it needs some form
of survivable penetrating ISR to reduce
pressure on the satellites. For example,
we used to have options for ISR like the
SR-71. Aircraft have the advantage of
flying unpredictable routes.
• The United States needs further
discussions of missile defenses and air
defenses. Do these matter? The Chinese
are definitely interested in air defense
and seem to be interested in missile
defense.
• The United States needs to consider
how it makes credible assurances to its
adversaries (a critical hard issue).
• The United States assumes, perhaps, that
it can deter the adversary by structuring
threats and posturing forces in certain
ways and that it can manage escalation
risks in certain ways, ignoring the fact
that other actors may be also contributing
to the situation. However, it may not be
able to constrain or restrain U.S. friends
and allies, who feel threatened and have
greater stakes in the region, even if they
notify us in advance.
The United States will need to reassure
them that it can take on the deterrence
and escalation management burdens to
protect them. It will need to establish
the credibility of its resolve, even in the
face of nuclear weapons. It may need
to structure deterrent threats against its
allies that will restrain them from taking
actions that would critically destabilize
the situation in a confrontation.
• The United States needs more discussion
of credible military and diplomatic
response options that might actually
be used in a crisis. This is intellectual
homework
for
thinking
about
deterrence. One of the best approaches
is through scenario analysis, some of
which presumably is conducted at the
classified level.
• The United States should address crossdomain deterrence challenges by using
rational strategic logic. Think about
these challenges as if the United States
49
Future Research
faced an anonymous adversary. How
would the United States deter actions in
domain A by threatening actions in
domain B? To address this challenge, it
needs to answer the following questions:
(1) What is the strategic context?
(2) What would be in U.S. interests?
(3) What would make a threat credible
or not credible? Rhetorical policy is
less relevant, and international norms
will not matter much once the shooting
starts.
• How do you deter a failing state or a
desperate state? This question has been
insufficiently examined.
• The United States needs to understand
the interplay of domestic and foreign
imperatives. The biggest threat to China
is internal. However, does China really
50
fear existential foreign threats or rather
those foreign threats that, in combination
with domestic instability, can become
existential
threats?
The
Chinese
Communist Party might not be able to
easily control the domestic situation
but could have an impact on deterring
the external threat using its nuclear and
conventional military forces.
• The United States needs to understand
China’s
perspective
on
multiple
deterrence. What if Russia is not working
with the United States, but after China
has deployed nuclear weapons at the
United States, does it have to worry about
having enough weapons left over to deter
or attack Russia and possibly other states?
How does China view a multipolar world
of multipolar deterrence?
7
FINAL THOUGHTS
Cross-domain deterrence, in reality, represents
new terminology and new applications of a
concept that is as old as nuclear deterrence
itself. After all, the first application of
deterrence in the Cold War was to prevent
Soviet conventional attacks against Western
Europe by invoking the fear of nuclear
response. This suggests that, by looking at
the historical applications of deterrence in
those situations in which multiple domains
were in play, analysts could develop lessons
applicable to emerging threats and future
scenarios.
We also are taken by the diversity of
answers to most questions. Clarity and
consensus on many cross-domain issues
clearly elude the U.S. strategic community.
It is unclear how to overcome this or whether
it is even possible to do so. Therefore, in
addition to historical analysis, it appears
that we might also learn from other states’
perspectives
and
policies
regarding
deterrence, cross-domain and otherwise.
Thus, a second potentially fruitful research
path would involve comparative analysis of
various states’ thinking on these questions.
Notwithstanding the general lack of
consensus, there was some agreement worth
noting. For example, few argued against
the notion that there is little likelihood of
a Chinese attack against U.S. space assets
outside the context of a broader conflict.
While we do not take issue with this judgment,
we are not confident in it to the point of
completely excluding from U.S. planning
the possibility of a Chinese surprise attack on
U.S. space assets. In fact, there is a perverse
dynamic at work here: those contingencies
that we do plan for decrease in likelihood
because of our planning, while those we do
not plan for increase in likelihood because of
our failure to plan for them. This suggests a
careful scrutiny of assumptions, explicit and
implicit, in cross-domain deterrence analysis.
Finally, this workshop was conducted at
what can be characterized as a conceptual
level. However, it became clear that
considering cross-domain issues at an
abstract level is limiting, as almost all crossdomain deterrence decisions are context
dependent. Thus, scenario analysis must be
an integral component of future research.
Moreover, as several participants have
noted, scenario analysis could greatly benefit
from the analysts trying to emulate Chinese
thinking.
As with many workshops, this one has
raised more questions than it has answered.
We hope that this compilation of the
perspectives expressed over the course of
this JHU/APL Cross-Domain Deterrence
Workshop will provide a useful point of
departure for other researchers as they pursue
these important questions.
51
Download